The Paradox of Adversity: New Left Party Survival and Collapse in Latin America

Transcription

1 The Paradox of Adversity: New Left Party Survival and Collapse in Latin America The Harvard community has made this article openly available. Please share how this access benefits you. Your story matters. Citation Accessed Citable Link Terms of Use Van Dyck, Brandon Philip The Paradox of Adversity: New Left Party Survival and Collapse in Latin America. Doctoral dissertation, Harvard University. June 20, :31:33 AM EDT This article was downloaded from Harvard University's DASH repository, and is made available under the terms and conditions applicable to Other Posted Material, as set forth at (Article begins on next page)

2 The Paradox of Adversity: New Left Party Survival and Collapse in Latin America A dissertation presented by Brandon Philip Van Dyck to The Department of Government in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the subject of Political Science Harvard University Cambridge, Massachusetts June 2013

3 c 2013 Brandon Van Dyck All rights reserved.

4 Dissertation Advisor: Jorge Domínguez Dissertation Advisor: Steven Levitsky Brandon Van Dyck The Paradox of Adversity: New Left Party Survival and Collapse in Latin America Abstract Political parties are the basic building blocks of representative democracy. They reduce information costs for voters, enhance executive accountability, and contribute to democratic governability by facilitating legislative organization and aggregating the interests of powerful societal groups. Yet we continue to know relatively little about the conditions under which strong parties form. The dominant theories of party-building are mostly based on historical studies of the United States and Western European countries, almost all of which developed stable party systems. Drawing on this literature, a segment of the early scholarship on party-building in third-wave democracies optimistically took party development for granted, assuming that parties would follow from democracy, cleavages, or certain electoral rules. Yet party-building outcomes in third-wave democracies fell short of scholars initial, optimistic expectations. In many third-wave polities, social cleavages, attempts at electoral engineering, and decades of democratic competition did not produce durable parties. On the other hand, in numerous third-wave democracies, new political parties did take root. What accounts for the variation in party-building outcomes observed across the developing world? More generally, under what conditions does party-building succeed? Drawing on evidence from fourteen months of interviews and archival research in Brazil, Mexico, Peru, and Argentina, this project provides the first systematic comparison of partybuilding success and failure on the Latin American new left. It argues that most new parties do not survive because they do not build strong organizations composed of committed iii

5 activists. As a result, they do not withstand early crises. Paradoxically, parties with strong organizations and committed activists are more likely to form under conditions of adversity. Office-seekers with low access to state resources and mass media have an incentive to do the difficult work of organization-building. Intense polarization and conflict (e.g., civil war, populist mobilization) generate committed activists by producing the higher causes that spur individuals and groups to collective action. New party-builders are more likely to experience this cluster of adverse conditions under authoritarian rule. iv

6 Table of Contents Tables and Figures Acknowledgments vi vii 1 The Paradox of Adversity: New Left Party Survival and Collapse in Latin America The Survival of Brazil s Workers Party 64 2 Building the PT: Adversity and Survival 66 3 Avoiding Schism in the Early PT: The Indispensability of Lula da Silva The Survival of Mexico s Party of the Democratic Revolution Building the PRD: Adversity and Survival Avoiding Schism in the Early PRD: The Indispensability of Cuauhtémoc Cárdenas Electoral Collapse: The Case of Argentina s FREPASO Collapse by Schism: The Case of Peru s United Left Conclusion 354 Bibliography 365 v

7 Tables and Figures Table 1.1: New Left Party Survival and Collapse in Latin America 8 Table 1.2: Authoritarian Origins and New Left Survival 55 Table 1.3: Organization and Activist Commitment 58 Table 1.4: Survival and Collapse 58 Figure 1.1: Conditions for Organizational Strength 29 Figure 1.2: Coattails and Internal Authority 42 Figure 1.3: The Role of Authoritarianism 46 Figure 1.4: Electoral History in Chamber PT, PRD, IU, FREPASO 50 Figure 2.1: The PT in the Chamber of Deputies (1982-present) 64 Figure 4.1: The PRD in the Chamber of Deputies (1991-present) 174 Figure 4.2: PRD Municipal Office Presence, Figure 4.3: PRD Municipal Office Presence, Figure 4.4: PRD Member/Population Ratio, Figure 6.1: FREPASO in the Chamber of Deputies 289 vi

8 Acknowledgments I owe an enormous debt of gratitude to my advisers, Jorge Domínguez, Steve Levitsky, Fran Hagopian, and Al Montero. They devoted many, many hours to this project and stuck with me through thick and thin (mostly thin). I am deeply grateful to all of them. I dedicate this dissertation to my parents, Madeline and Phil Van Dyck. vii

9 Chapter 1 The Paradox of Adversity: New Left Party Party Survival and Collapse in Latin America Political scientists rarely agree, but on the importance of political parties, they speak in near unison. Parties play an indispensable role in representative democracy. 1 They reduce information costs for voters, 2 extend politicians time horizons, 3 and contribute to democratic governability and stability by socializing elites, aggregating the interests of powerful societal groups, facilitating legislative organization, and enhancing executive accountability. 4 Yet parties remain weak in much of the developing world. Over the last two decades, established parties have declined or collapsed across Africa, Latin America, Asia, and the post-soviet world. Moreover, despite decades of competitive elections and repeated efforts at electoral reform, relatively few new parties and party systems have taken root. Strong 1 [M]odern democracy, writes Schattschneider (1942), is unthinkable save in terms of political parties (1). 2 By reducing information costs for voters, parties increase political participation, a central value within normative democratic theory. On Downs (1957) classic account, voting exacts costs to the individual, not only in money and transportation time, but also in the time and effort required to form a preference between candidates. Because people are busy and/or relatively uninterested in politics, they are unlikely to acquire sufficient candidate information in the absence of information shortcuts. Downs argues that parties, by representing simple principles and policy stances, make it easier for ordinary citizens to form preferences between numerous candidates. Hence, parties increase the likelihood that ordinary citizens will vote and become politically active. 3 By extending politicians time horizons, parties improve long-term policy outcomes. Parties have instrumental value for office-seekers, providing the money and organizational resources necessary to run successful campaigns, and in some cases delivering a significant portion of automatic votes from partisans in the electorate (Aldrich 1995: 24-5). Yet parties, viewed as unitary actors, are more likely than individual politicians to have long time horizons and national as opposed to local or regional goals, and hence to stake out positions in favor of policies and ideas with long-term, universal benefits for national populations (Levitsky and Cameron 2003: 3). 4 Parties help protect the interests of powerful societal groups, whether elites (Gibson 1996) or mass-based civil society organizations (Collier and Collier 1991), reducing the likelihood of mass praetorianism (Huntington 1968). Parties make legislative debate more simple and efficient, unite the positions of large groups of legislators, and sometimes unite the positions of presidents and groups of legislators, all of which increase democratic governability (Mainwaring and Scully 1995, Mainwaring 1999, Levitsky and Cameron 2003). They act as a check on presidential power, socialize elites, and when firmly rooted in the electorate reduce the likelihood that political independents will gain power. In all these ways, they help check or prevent the emergence of major politicians who lack governing experience and a strong commitment to democratic institutions. 1

10 new party systems have emerged in some countries, 5 but in many others they have not, 6 and in a few countries party systems have decomposed entirely. 7 Party weakness has destabilized democracy. Where party systems have collapsed and not been rebuilt, democracies have often fallen into crisis (e.g., Russia). 8 In contrast, where successful party-building has occurred, democracies have typically become consolidated (e.g., Brazil). 9 Despite the clear importance of parties, we continue to know relatively little about the conditions under which strong parties form. The dominant theories of party-building are mostly based on historical studies of the United States and Western European countries, almost all of which developed stable party systems. 10 These works explore how electoral institutions (Duverger 1954; Lijphart 1994), social cleavages (Lipset and Rokkan 1967), and access to patronage (Shefter 1994) shape emerging party systems, leaving aside the more 5 E.g., Brazil, El Salvador, Ghana, Taiwan. 6 E.g., Ecuador, Kenya, Russia, Venezuela. 7 E.g., Peru, Guatemala. Sánchez (2009) calls the Guatemalan party system a party non-system. 8 Also Venezuela, Peru, Ecuador. In 1990s Venezuela and 2000s Ecuador, party system collapse facilitated the rise of charismatic, media-savvy neopopulists who won presidential elections, campaigning on their dissociation from the party establishment and capitalizing on the popular perception of political parties as ineffectual and exclusionary (Hugo Chávez in Venezuela, 1998; Rafael Correa in Ecuador, 2006). Upon entering the presidency, both leaders concentrated power in the face of weak partisan opposition and ultimately dismantled core democratic institutions (Levitsky and Loxton n.d.). Chapters Two to Three and Chapter Seven of this book will suggest that in Brazil and Peru, the contrasting fates of new left parties/coalitions determined subsequent democratic outcomes. Brazil s Workers Party and Peru s United Left coalition (IU) both formed in The PT developed within a highly inchoate party system, the IU in a semi-institutionalized system with two major national parties (APRA, AP). Both the PT and IU had popular mass bases, promoted distinctive programs, and quickly came to anchor the left poles of their national party systems. The PT survived its first decade, after which the Brazilian party system progressed from inchoate to semi-institutionalized, and Brazil became a consolidated democracy. Hagopian (forthcoming) identifies the survival of the PT as crucial for the stabilization of Brazil s party system. The IU collapsed in 1990, and Peru subsequently experienced party system disintegration and democratic breakdown. Lynch (2000) identifies the IU s failure by schism as a precondition for the election of antisystem outsider, Alberto Fujimori, and the resulting collapse of Peru s party system and democracy. 9 Also Ghana, Taiwan. 10 Duverger 1954; Lipset and Rokkan 1967; Sartori 1976; Panebianco 1988; Kitschelt 1989; Lijphart 1994; Shefter 1994; Aldrich 1995; Kalyvas

11 fundamental question why parties take root in the first place. Some of the classic scholarship posits, or rests on the premise, that basic conditions such as electoral competition, social cleavages, and certain formal institutions should give rise to strong parties (e.g., Aldrich 1995). Drawing on this literature, a segment of the early scholarship on party-building in third-wave democracies optimistically took party development for granted, assuming that parties would follow from democracy, cleavages, or certain electoral rules (e.g., Brader and Tucker 2001). Yet party-building outcomes in third-wave democracies fell short of scholars initial, optimistic expectations. In many third-wave polities, social cleavages, attempts at electoral engineering, and decades of democratic competition did not produce durable parties. The perceived failure of third-wave party-building gave rise to a reverse, pessimistic trend in the literature on party development. Scholars began to identify contemporary impediments to party-building, emphasizing heightened social fragmentation (Roberts 1998), programmatic convergence around free markets and democracy (Roberts 1998), elite access to mass media (Mainwaring and Zoco 2007), and elite access to party substitutes such as state resources and business conglomerates (Hale 2006). 11 Yet neither the optimistic view that parties should emerge under democracy, nor the pessimistic view that the era of party-building has passed, corresponds to the facts. The thirdwave party-building record presents considerable variation. On the one hand, in many new democracies, politicians have not invested in parties, despite their supposed utility, and parties have not followed from social cleavages or electoral engineering. 12 On the other hand, 11 See also Barndt (n.d.). 12 Politicians have passed electoral reforms to create incentives for party-building in a range of third-wave democracies, but these reforms have done little to produce strong parties. In many cases, parties have weakened following the implementation of such reforms (e.g., Dargent and Mu noz n.d.) and some of the most successful new parties in recent decades seem to have taken root despite prevailing electoral institutions (Brazil s PT, PSDB, and PFL. 3

12 in numerous third-wave democracies, new political parties have taken root. What accounts for the variation in party-building outcomes observed across the developing world? More generally, under what conditions does party-building succeed? Defining new party survival The book defines political party as a group of individuals contesting elected office under a common label. 13 Yet the central theoretical question of the book Under what conditions does party-building succeed? 14 does not concern all attempts at party-building. It only concerns major new parties and coalitions: those that, for at least one shining moment, become serious electoral contenders at the national level. 15 To be considered a serious national contender, new parties do not need to win national elections, presidential or legislative. They should, however, be in the running. More importantly, they must win enough elections at the congressional level, and possibly at the subnational executive and legislative levels to play major roles in their country s politics. New party survival denotes medium-term electoral significance at the national level. 16 The book examines why, among parties that seriously contend for national power, some endure on the national stage, while others collapse. The book equates new party survival with survival of the formative phase (Panebianco 1988), roughly the first decade of a party s existence. Parties that never achieve national electoral significance do not receive attention, independently of whether they endure the formative phase as marginal actors. 13 Epstein (1967); Chhibber and Kollman (2004); Hicken (2009). 14 Alternatively: Under what conditions do new parties survive? 15 Although some coalitions do not come close to cohering, as coalition elites make no attempts to unify the relevant constituent parties, other coalitions persist across multiple elections, and major elites actively pursue unification under a single party label (e.g., Peru s Izquierda Unida). The book treats the latter type of coalition as a party, recognizing that party may be a slight misnomer. 16 For a similar definition, see Levitsky, Loxton, and Van Dyck (n.d.). 4

13 As a final qualification, surviving party means surviving party, not surviving politician. Parties that depend on a charismatic national leader for down-ticket electoral success tend to collapse or become marginal as soon as the candidate in question exits the political stage. Thus, personalistic vehicles even long-lasting ones (e.g., Bolivia s ADN under Hugo Banzer) and previously institutionalized ones (e.g., Peru s APRA under Alán García) 17 are not successful parties in their own right. The book s central theoretical question can now be restated as follows: Why do some new parties/coalitions that become serious national contenders survive the formative phase, while others collapse after their initial success? New left party survival and collapse in Latin America Latin America s new left parties and coalitions offer an ideal setting for examining the roots of party-building success, both in general and in third-wave polities specifically. These new left parties (tabulated below) emerged during the third wave, a period in which the vast majority of Latin American countries shifted from military rule to democracy. At the end of the 1970s, military regimes ruled all but a handful of Latin American countries, 18 but in a period of several years, nearly all of these regimes fell, some quickly collapsing (e.g., 17 In her analysis of party rebuilding after system collapse, Cyr (n.d.) defines party revival in terms of presidential performance, thus scoring Peru s APRA as a successful comeback due to Alán García s 2006 presidential election victory. APRA s electoral significance, however, depends entirely on García. In 2006, APRA won 36 of 120 seats (thirty percent) in Peru s unicameral legislature. In 2011, APRA failed to produce a successor candidate to García, and the party elected four of 130 seats (three percent). These facts indicate that APRA is no longer a successful party in its own right. 18 In the 1960s and 1970s, Latin American militaries, backed by economic elites and other informal civilian allies, toppled democratically elected governments and assumed power across the region. Latin America s military regimes took diverse forms. Some military governments pursued statist and redistributive policies (e.g., Peru under Velasco; Brazil s BA regime), others adopted radical neoliberal reforms (e.g., Chile s Pinochet/BA regime), and many sought to neutralize the perceived Marxist security threat through hard repression (e.g., all Southern Cone BA regimes). 5

14 Argentina s Dirty War regime), others gradually liberalizing (e.g., Brazil s military regime). 19 Several established opposition parties capitalized on the breakdown of authoritarian rule by contesting national power under new democracy. 20 Simultaneously, a large number of new opposition parties from the left, center, and right emerged on the scene and became serious contenders for national power. Some of these new parties, from inception, shared the electoral arena with established parties (e.g., Mexico s PRD). 21 Others were born into systems with no established parties (e.g., Brazil s PT). 22 A similar process unfolded in the few Latin American countries that had not experienced democratic breakdown during the 1960s and 1970s (e.g. Colombia, Venezuela). In these countries, established parties that had dominated electoral politics for decades 23 began to face serious electoral competition from newer parties. 24 Scholars conventionally designate 1978 as the date of the third wave s onset in Latin America. 25 Between 1978 and the mid-1990s, hundreds of political parties emerged and competed in national elections in the region. A few dozen became serious national contenders, and among these, a dozen have survived to the present. 26 Of the twelve survivors, six belong 19 The decline of the Soviet Union undermined military regimes on the left and right. The failure of the Soviet model discredited the statism favored by left military governments, while the erosion of Soviet power and influence removed a central justification for right-wing military rule (i.e., the perceived domestic communist threat). These challenges to authoritarian legitimacy, in combination with a region-wide debt crisis and increased pressure for democracy from Western governments and transnational advocacy networks (Keck and Sikkink 1998), gave rise to Latin America s third wave. 20 E.g., Argentina s PJ, Chile s Socialists, Mexico s PAN, Peru s APRA, Uruguay s Blancos and Colorados. 21 E.g., Argentina s UCéDé and FREPASO, Chile s UDI, RN, and PPD, Mexico s PRD, Paraguay s PEN, Peru s IU and ML, Uruguay s FA. 22 E.g., Bolivia s MAS, Brazil s PT, PFL, PSDB, El Salvador s ARENA and FMLN, Guatemala s FRG, FDNG, and MAS, Nicaragua s FSLN. 23 E.g., Venezuela s AD and COPEI, Colombia s Liberals and Conservatives. 24 E.g., Venezuela s LCR and MVR/PSUV, Colombia s AD-M19 and Partido de la U. 25 The third wave is conventionally dated from the mid-1970s onward globally (Huntington 1993), but from 1978 onward in Latin America (Hagopian and Mainwaring 2005). 26 Bolivia s Movement toward Socialism (MAS); three Brazilian parties, the Workers Party (PT), Brazilian Social Democracy Party (PSDB), and Liberal Front Party (PFL); three Chilean parties, the Independent 6

15 to the left, 27 placing central programmatic emphasis on the reduction of poverty and inequality in the economic, social, and political spheres: Bolivia s MAS, Brazil s PT, Chile s PPD, El Salvador s FMLN, Mexico s PRD, and Nicaragua s FSLN. 28 An equal number of new left parties and coalitions collapsed after becoming serious national contenders during the formative phase: Argentina s FREPASO, Colombia s M-19, Guatemala s MAS and FDNG, Paraguay s PEN, and Peru s IU. Table One provides the full sample of new left parties/coalitions born between 1978 and 1995 that received ten percent of the vote in at least one congressional election. 29 Those that have maintained the ten-percent threshold for five consecutive elections are treated as cases of survival, the rest as cases of collapse. The use of legislative electoral performance, not presidential performance, 30 as an indicator of success serves to distinguish parties from personalistic vehicles, at least to the extent possible. 31 The study s operational cutoff points (ten percent of vote share, five consecutive elections) constitute a reasonable middle ground between higher and lower alternatives that Democratic Union (UDI), National Renewal (RN), and Party for Democracy (PPD); two El Salvadoran parties, the Nationalist Republican Alliance (ARENA) and Farabundo Martí National Liberation Front (FMLN); Mexico s Party of the Democratic Revolution (PRD); Nicaragua s Sandinista National Liberation Front (FSLN); and Panama s Democratic Revolutionary Party (PRD) (Levitsky, Loxton, and Van Dyck n.d.). 27 (Definition from Levitsky and Roberts (2010). 28 Some studies highlight distinctions between Latin America s major new left parties and coalitions. Roberts (1998) distinguishes between the populist and organic left; Castañeda (2006) between two subspecies of contemporary Latin American left parties and leaders, one modern, open-minded, reformist, and internationalist (e.g., Brazil s PT), the other nationalist, strident, and close-minded (e.g., Venezuela s MVR/PSUV); Weyland, Madrid, and Hunter (2010) between the moderate and radical (or statist) left; and Levitsky and Roberts (2010) between left parties characterized by concentrated and dispersed authority. 29 No party born after 1995, including Venezuela s MVR/PSUV, has competed in enough successive congressional elections to qualify as a success. 30 Legislative election results may reflect coattail effects, and even if they do not, independently successful party legislators may have a fragile association with the party as such, participating out of loyalty to the leader, or because of the electoral incentives that the leader provides. 31 Some new left survivors (e.g., Bolivia s MAS) still have not experienced a leadership transition, in contrast to others such as Brazil s PT and Mexico s PRD. If a party in the sample had maintained the ten-percent threshold for five consecutive elections but not survived the exit of its main leader (e.g., Bolivia s out-ofsample ADN, led by Hugo Banzer), it would have been scored as a failure. 7

16 do not cohere with our intuitive case-level assessments. For example, Middlebrook (2000: 4) operationalizes party success as twenty percent of the vote in two consecutive congressional or presidential elections, but on this measure, Argentina s meteoric new left party, FREPASO, would qualify as a success even though it collapsed within a decade of its creation and just several years after becoming a serious national contender. On the other hand, if one lowered the threshold significantly below ten percent, one would have to treat niche or regional parties as major contenders or even full-fledged successes, as some such parties last for very long periods. Table 1.1: New Left Party Survival and Collapse in Latin America Party Country Founded Outcome FSLN Nicaragua 1979 Survival IU Peru 1980 Collapse PT Brazil 1982 Survival PPD Chile 1987 Survival AD-M19 Colombia 1989 Collapse PRD Mexico 1989 Survival PEN Paraguay 1991 Collapse FMLN El Salvador 1992 Survival MAS Guatemala 1993 Collapse FREPASO Argentina 1994 Collapse FDNG Guatemala 1995 Collapse MAS Bolivia 1995 Survival 8

17 The conundrum of new left success The record of new left party-building in Latin America presents a conundrum. After the third wave s onset, six new left-wing parties emerged and became institutionalized as major national contenders. Only two solidly right-wing new parties achieved the same: El Salvador s ARENA (est. 1981) and Chile s UDI (est. 1983). 32 Judging from the regional literature on the new left, which strikes a decidedly pessimistic tone, Latin America s new left parties should have fared much worse. The 1980s produced a set of challenges for left-wing party-building in Latin America. The region s debt crisis, the implosion of its statist/protectionist economic model, and the simultaneous decline and collapse of the Soviet Union, gave rise to a policy consensus around free markets, or neoliberalism, within Latin America. 33 After the early 1980s, the region s left parties paid an electoral price for promoting left economic policies, or for implementing them once in power. 34 The neoliberal consensus thus impeded left-wing programmatic differenti- 32 See Loxton n.d. 33 Neoliberalism denotes a set of economic policies, primarily orthodox stabilization and structural adjustment, which are usually associated with the political right and roughly conform to the Washington Consensus (Williamson 1989). Orthodox stabilization encompasses fiscal austerity, achieved mainly through government spending cuts, and anti-inflationary monetary policy, which typically features high interest rates on inter-bank lending. Structural adjustment entails, first, increasing private-sector exposure to international market competition by eliminating protective tariffs and lifting restrictions on foreign direct investment (FDI) and, second, reducing the state s role in production by privatizing and deregulating industries. 34 Prior to the third wave and neoliberal turn, authoritarian regimes ruled most of Latin America, and large segments of the political class and electorate viewed socialism as a potential alternative to the Western capitalist model. These conditions allowed for the emergence of political parties based on fundamental economic and regime positions, such as opposition to capitalism and support for democracy. In the 1980s and 1990s, however, with government debt to Western lenders skyrocketing, and economic statism and protectionism losing favor as a growth strategy, most Latin American governments turned to neoliberalism. In several short years, governments across the region reversed decades of statist, protectionist, and often highly inflationary economic policies and adopted the major economic reforms privatization, deregulation, trade liberalization, fiscal austerity, tight monetarism favored by Western policymakers and intellectuals, promoted by Western governments, necessitated by global market shifts, and in some cases explicitly demanded by Western lenders like the IMF (Stallings 1992). As external pressures mounted and growing numbers of left-leaning voters, technocrats, and political elites came to view hallmarks of the left s traditional platform as infeasible or undesirable (e.g., land reform, the socialization of core industries), US- 9

18 ation, programmatic consistency, and hence brand development (Lupu forthcoming). 35 The discrediting of Marxism generated an identity crisis, and also produced new ideological fault lines, within the Latin American left. Forced to grapple with the defeat of revolutionary leftist ideas, many left-wing networks and organizations in Latin America reformed, or renovated, their core beliefs and policy stances, accepting the basic features of the free market and advocating a range of social and economic policies designed to smooth capitalism s rough edges. 36 Yet revolutionary leftism did not disappear. To different degrees in different countries, the radical left remained a potent force, and internal divisions between radicals and moderates posed a threat to the basic ideological integrity of new left parties and coalitions (Roberts 1998). The decline of organized labor and rise of the informal sector 37 made popular incorporation and coalition-building more difficult for left parties. As Latin American governments abandoned protectionism, domestic industries folded and shrank, and industrial trade unions, on which left parties had traditionally depended for votes and resources, folded and shrank with them. 38 The shedding of manufacturing jobs, coupled with the general unemployment style capitalism became the only game in town (Linz and Stepan 1996). In the new world order of the 1980s and 1990s, however, left parties did not have authoritarian regimes to oppose and could no longer take economic and social policy cues from the Soviet Union, as many had done for decades. Anti-capitalist platforms tended to have limited electoral appeal, and even where left parties managed to gain electoral traction by opposing capitalism, once in government they confronted insurmountable obstacles to sweeping economic reform. 35 In addition, the recessions, budget crises, and inflationary spirals of the 1980s increased the likelihood of weak performance for left and right governments alike (Levitsky, Loxton, and Van Dyck n.d.). 36 E.g., selective nationalization, regulation, a stronger public safety net. 37 The informal sector encompasses all economic activity excluding illegal trafficking that takes place outside the regulatory norms of the state (e.g., taxes, licensing, labor regulations, zoning ordinances) (Cross 1998: ). 38 By the late 1980s and 1990s, organized labor had become too small a constituency to ground a successful, national party-building project. Brazil s Workers Party (PT) is no exception. Unions gave the PT strength in certain regions of Brazil (particularly São Paulo), but the PT depended on other organized constituencies elsewhere. See Chapter Two. Moreover, the new left often could not count on the support of organized labor, given unions longstanding ties to established parties, especially those with classic populist origins (Murillo 2001; Levitsky 2003). 10

19 spikes of the debt crisis, caused an explosion of the informal sector. 39 In order to reach this ballooning, atomized social class, 40 new left parties had to make difficult programmatic adjustments. Traditional hallmarks of the left platform such as wage indexation did not resonate with informal workers. Increasingly, left electoral success would require segmented appeals, with politicians and parties attempting to build heterogeneous coalitions of formal and informal workers, as well as rural laborers and middle-class groups. 41 This segmentation would necessitate a careful balancing of interests between distinct, potentially conflicting voting blocs. 42 Given the formidable obstacles to left-wing party-building in contemporary Latin America, the success of the new left, relative to the new right, merits close examination. Six major new left parties took root, and not because the difficulties highlighted by scholars did not obtain in their cases. On the contrary, most of Latin America s new left survivors faced the full range of difficulties associated with the debt crisis, decline of communism, and neoliberal turn. 43 Mirroring the classic literature on party system development, current literature on new left 39 Informal workers became the largest social class in many Latin American countries, including Brazil and Mexico, the two most populous countries in the region (Haber 2008). 40 Informal workers tended to exhibit weaker collective identities than their counterparts in organized labor, and the geographical dispersion of informal workers often prevented the creation of sectoral and cross-sector associations (Roberts 1998). Thus, although specific sectors of the informal working class occasionally became unionized (e.g., street vendors in Mexico City (Cross 1998)), most did not do so, and there was virtually no cross-sectoral unionization. For a discussion of informalization and collective action, see Roberts (1998: 63-73). 41 Although left parties in Latin America had long possessed fluid, diverse constituencies, especially relative to their Western European labor-based counterparts, neoliberal-era deindustrialization amplified these characteristics. The downsizing of organized labor and the rapid growth of the heterogeneous informal sector fragmented the left s potential activist and voter bases. Whereas classic populist parties in Latin America had relied on a central support base (organized labor), the new left could not be defined by its appeal to a core social constituency (Levitsky and Roberts 2010). 42 See Luna (n.d.); Bruhn (1998). 43 I.e., Brazil s PT, El Salvador s FMLN, Mexico s PRD, Nicaragua s FSLN (see Chapters Two through Five and Chapter 8). Moreover, numerous new left parties that collapsed arguably faced less extreme versions of these difficulties (e.g., Argentina s FREPASO, Colombia s AD-M19). 11

20 parties in Latin America focuses disproportionately on successful cases, or cases of survival. While book-length analyses of new left survivors number in the hundreds, 44 book-length analyses of new left failures number in the single digits. 45 In addition, most books and articles on failed cases were published before the parties in question had fully collapsed 46 or even begun to collapse. 47 These non-retrospective analyses typically assumed, explicitly or implicitly, that the relevant parties would survive. 48 Many studies of new left survivors highlight variables (e.g., institutional variables) that also obtain for failed new left parties and hence cannot account for variation. 49 Other analyses of new left survivors leave out or bury variables that do help explain variation, such as mass media access 50 and the party leader s internal network ties. 51 Without giving attention to cases of party-building failure, we cannot understand the causes of party-building success E.g., Keck (1992); Coppedge (1994); Bruhn (1998); Roberts (1998); Loaeza (1999); Borjas (2003); Wood (2003); Chavez and Goldfrank (2004); Van Cott (2005); Castañeda and Morales (2007); Greene (2007); Ortega (2008); Wuhs (2008); Hunter (2010); Kitschelt et al. (2010); Levitsky and Roberts (2010); Martínez González (2010); Ribeiro (2010); Goldfrank (2011); Secco (2011); Mossige (2013); Rodríguez (n.d.). 45 Retrospective book-length studies of failed cases include Herrera (2002) and Adrianzén ed. (2011) on Peru s IU and Valenzuela (2003) on Argentina s FREPASO. 46 E.g., López-Maya (1997) on Venezuela s LCR; Boudon 1997 on the LCR and Colombia s AD M E.g., Novaro and Palermo 1998 on Argentina s FREPASO. 48 López-Maya 1997, Novaro and Palermo Ribeiro (2010), for example, attributes the early PT s cohesion, in part, to the implementation of full proportional representation in the National Executive Committee roughly a decade after the party s founding. Yet Peru s United Left coalition (IU) split apart despite having full proportional representation in the National Leadership Council (CDN) from its founding onward. For details, see Chapters Three and Seven. 50 See Chapters Two through Six. 51 See Chapters Three, Five, and Seven. 52 In the last two decades, political scientists have devoted increased attention to methodological problems associated with small-n, qualitative research (Munck 1998). One strand of this literature focuses on theory development (Mahoney 2005), and within this strand, authors have focused in particular on the problem of poor case selection. Critics argue that small-n qualitative researchers are prone to selection bias and, in particular, to selecting on the dependent variable, or examining cases with the same value or similar values on the outcome of interest (King, Keohane, and Verba 1994; Achen and Snidal 1989; Geddes 2003). From the perspective of large-n, statistical research, selection on the dependent variable prevents researchers from identifying the independent variables and interactions that vary systematically with change in the value of the dependent variable. This limits the extent to which qualitative researchers, in cross-case studies, can make valid causal inferences. According to Barbara Geddes, for example, selection on the dependent 12

21 Moreover, the reliance on single- and two-case studies, as opposed to medium-n studies, has led scholars to provide descriptive, laundry-list accounts, or to emphasize explanatory factors that do not travel across successful cases (e.g., electoral laws, 53 internal democracy 54 ). Small-n studies are also biased toward voluntarism, often assigning inordinate weight to the contingent choices of party leaders. 55 In short, selection on the dependent variable has caused new left scholars to present unpersuasive arguments or underemphasize key explanatory variables. In turn, small-n research designs have biased analyses toward voluntarism and, more generally, prevented scholars from identifying explanatory variables common across cases of success. Theories of party-building In recent years, a growing group of scholars has seized on empirical variation in third-wave party systems, or reexamined variation in the traditional party systems of the West, in order variable has tainted the findings of many classic cross-case, small-n, qualitative studies on topics such as economic development and revolution (Geddes 2003). In response to this critique, some have argued that the value of qualitative research (e.g., the in-depth tracing of complex causal processes) does not require variation on the dependent variable (Hall 2007). Others have promoted qualitative methods intended to solve the problem of poor case selection. For a review of these methods, see Mahoney (2005). 53 Scholars of Brazil s PT and Mexico s PRD, for example, have emphasized the role of electoral institutions, particularly legal registration requirements, in the successful organizational development of both parties (Keck 1992; Ribeiro 2010; Rodríguez n.d.). Keck (1992) writes, for example, that the legal requirements of the new party law passed in 1979 had a crucial impact on the structure of the Workers Party and on its efforts to become an internally democratic mass membership party (86). Yet other new left successes (e.g., Nicaragua s FSLN, El Salvador s FMLN, Bolivia s MAS), and even some failures (Peru s IU, Venezuela s LCR), built strong organizations in the absence of comparable registration requirements. 54 PT scholars and members frequently identify the party s vibrant, base-level participatory democracy as crucial to the party s early success. Most new left successes in Latin America, did not have comparable institutions of participation and debate at the base level during early development. For details, see Chapters Two through Five and Chapter Eight. 55 Scholars have partially attributed the success of Mexico s PRD to the charismatic leadership of Cuauhtémoc Cárdenas (Borjas 2003) and the success of Brazil s PT to leader Lula da Silva s charismatic leadership, promotion of internal democracy, and emphasis on organization-building (Hilgers 2008; Hunter 2010). Others have retrospectively linked FREPASO s collapse to Carlos Chacho Álvarez s insistence on rapid electoral growth (Abal Medina 2009), the AD-M19 s failure to party leader Antonio Navarro s inattention to organization-building (Boudon 2001; see also García et al. 2008), the fracturing of Venezuela s LCR to founder Alfredo Maneiro s unexpected death in 1982 (López-Maya 1997), and the disintegration of the IU, in part, to coalition leader Alfonso Barrantes s aloofness (Roberts 1998). 13

22 to generate and test new theories of party-building. 56 To what extent do these theories, almost all based on non-latin American cases, shed light on new left party survival and collapse in Latin America? Institutions and party-building Literature examining the impact of formal institutions on party development has multiplied in recent decades. Several key perspectives have emerged from this literature. One strand argues that low legal barriers to entry facilitate the creation, or emergence, of new parties (Moser 1999, 2001; Van Cott 2005). On this account, low thresholds for party registry, ballot access, and electoral representation facilitate party formation by enabling new parties to establish a toehold in the political system. 57 Another strand links low party system fragmentation to centralizing political institutions. Chhibber and Kollman (1998, 2004) argue that in vertically centralized polities, 58 voters do not reward subnational governments for policy outcomes, and subnational elites thus aggregate nationally on pain of electoral marginality (e.g., the US). 59 Building on Chhibber and Kollman, Allen Hicken argues that party system fragmentation decreases in horizontally centralized political systems, where elites have incentives to coordinate with each other across subnational units. Hicken (2006, 2009) attributes the recent, precipitous decrease in 56 Kitschelt (1989); Kalyvas (1996); Chhibber and Kollman (2004); Hale (2006); Hicken (2009); Hanson (2010); Samuels and Shugart (2010); LeBas (2011); Arriola (2013); Riedl (forthcoming). 57 Moser (1999, 2001) finds, for example, that proportional representation systems ease party-building by blocking outsider candidates and hence making it easier for new parties to elect a critical mass of national legislators. Van Cott (2005) argues that low electoral, financial, and infrastructural barriers have facilitated ethnic party formation in Latin American countries such as Bolivia and Ecuador, and that high thresholds have prevented ethnic party formation in Peru. 58 I.e., Weakly federalized polities. 59 Chhibber and Kollman (1998, 2004) argue that regional parties tend to proliferate in more federal (vertically decentralized) polities such as contemporary India and Canada. Other accounts suggest, however, that by giving subnational governments policy autonomy and control of state resources, federal systems facilitate party-building by helping nascent parties consolidate at the local and regional levels and scale up from there (Holland n.d.). 14

23 Thailand s effective number of national parties to the country s 1997 constitutional reforms, which imposed strict penalties for party-switching and created incentives for parliamentary coordination through the implementation of mixed-member proportional representation. 60 He contrasts Thailand with the Philippines, where the party system fragmented after President Marcos 1986 ouster and the passage of reforms weakening the national executive (e.g., the presidential reelection ban). 61 A final strand of the institutionalist literature, represented by Samuels and Shugart (2010), suggests that presidential systems impede party-building. According to Samuels and Shugart, presidential systems weaken parties by giving substantial patronage access and policy influence to presidents, who are not subject to removal by the legislature or, it follows, from their own party (assuming they belong to one). Parties in presidential systems thus tend to be less disciplined, as the executive and legislative wings may clash, and also less autonomous, given their inability to hold the executive branch accountable. For these reasons, presidential systems reduce elite incentives for party-building. Their merits notwithstanding, the above institutionalist arguments possess two main limitations for the purposes of the present analysis. First, most of them do not strictly concern new party survival, instead examining the causes of new party emergence (Moser 1999, 2001; Van Cott 2005), or of variation in the number of national parties (Chhibber and Kollman 2004; Hicken 2009). Second, and more importantly, in contemporary Latin America, the key variables highlighted in existing institutionalist theories either do not vary across diverse 60 Thailand s party system fragmentation decreased precipitously in the late 1990s and 2000s. The PTP (formerly PPP) was born in 1998 and is now the country s leading national party. In the 1997 constitution, party-switching penalties included laws prohibiting cabinet members from returning to the legislature and laws lengthening the period required to join a new party. The mixed-member PR system, modeled on Germany s, stipulated that 4/5 of lower-house candidates run in single-member districts. 61 Hicken s analysis centers on the presidential reelection ban, which discouraged presidents from investing in parties and, by eliminating incumbency, weakened the incentive for internal and external challengers to coalesce around the president or a single opposition candidate. 15

24 party systems, or appear to distribute randomly across them. Every new left party in Latin America, successful or failed, developed in a presidential system with full or mixed-member proportional representation. 62 New left parties have succeeded in both federal (e.g., Brazil, Mexico) and unitary systems (e.g., Chile, El Salvador, Uruguay), in systems with both high (e.g., Brazil, Mexico) and relatively low barriers to entry (e.g., Bolivia), and in systems with electoral rules considered inimical to horizontal centralization (e.g., Brazil). 63 Equally, new left parties have failed in a wide range of institutional contexts (e.g., Argentina, Colombia, Peru, Guatemala). 64 Access to resources and party-building A second approach attributes party-building outcomes to variation in access to resources (i.e., patronage and finance). Most scholars within this approach hold, straightforwardly, that access to patronage, public finance, or private finance facilitates party-building. Shefter (1994) shows, for example, that the founders of the United States Democratic and Republican Parties used state patronage to mobilize their supporters and construct national party organizations. Chandra (2004) finds that India s ethnic parties succeeded where they had enough state patronage at their disposal to attract elites within their ethnic group. Bruhn (n.d.) identifies an association between public finance and party system institutionalization in Latin America. 65 Along complementary lines, several scholars have argued that low ac- 62 Mexico s PRD and Venezuela s LCR, a shadow case covered in the conclusion, developed under mixedmember PR. The rest developed under full PR. 63 On the latter point, see Mainwaring (1999). 64 It should also be noted that, according to a sizable body of scholarship, electoral institutions in thirdwave democracies are often endogenous to deeper, structural factors such as authoritarian incumbent strength during the transition. See, for example, Stepan (1988); Garretón (1995); Linz and Stepan (1996); Ahmed (2010); Riedl (forthcoming). Riedl (forthcoming) finds that during democratic transitions in sub- Saharan Africa, strong authoritarian incumbents (e.g., in Ghana, Senegal) imposed electoral rules that [forced]...aggregation into fewer effective parties, and polarization into discrete incumbent and opposition camps (viii). Where outgoing regimes were too weak to impose such rules (e.g., Zambia, Benin), party systems remained fragmented. 65 Bruhn also argues, more specifically, that access to Mexico s generous public party fund enabled the new 16

25 cess to state resources impedes party-building (e.g., Fish 2005). 66 Arriola (2013) focuses on private finance, arguing that financial liberalization has facilitated multiethnic coalitionbuilding in sub-saharan Africa by giving opposition politicians access to business capital and, consequently, the capacity to purchase cross-ethnic endorsements (183). 67 Others hold, however, that access to resources, especially state resources, actually impedes party-building. In his analysis of party non-formation in post-soviet Russia, Hale (2006) finds that federal and provincial politicians have systematically used the state and, to a lesser extent, business as a substitute for parties. Hale observes that investing in parties would require Russian political elites to sacrifice resources and autonomy without the guarantee of a worthwhile personal return, electorally or materially. Rather than investing in traditional party organization, these elites have sought to maintain their electoral clout by mobilizing their bureaucratic fiefdoms for electoral campaigns and tapping both government agencies and corporations for money and candidate allies. 68 Hale s analysis appears to travel to other inchoate third-wave party systems (e.g., Belarus, Ukraine, Peru), where politicians in power have eschewed party-building once in office, instead selecting candidates from within the government apparatus and deploying government workers to perform the functions often reserved for party activists. 69 left PRD to finance national campaigns and sustain a national party organization. 66 Fish (2005) argues that robust political parties did not emerge in post-soviet Russia, in part, because the state monopolized property, employment, and industry, and would-be opposition parties thus faced an overwhelming material disadvantage. Dargent and Muñoz (n.d.) identify reforms depriving regional barons of access to patronage as a key impediment to party-building in contemporary Colombia. 67 Also see Luna (2010), who finds that access to private finance enabled Chile s new right UDI to engage in private clientelism and thus to shore up support among the informal urban poor during early development. 68 Along similar lines, Reuter and Remington (2009) observe that in 1990s Russia, governors who controlled powerful provincial machines refused to give up their resources and autonomy to a national party with an uncertain future. 69 E.g., Presidents Leonid Kuchma of Ukraine, Alexander Lukashenko of Belarus, Alberto Fujimori of Peru, and Ferdinand Marcos of the Philippines (Smith 2005: 447), and the leaders of the West African partystates (Ghana, the Ivory Coast, Guinea, Mali) (Zolberg 1985: 94). In Peru during the 1990s, Alberto Fujimori made heavy use of his allies within the national security establishment, deploying the armed 17

26 Although the above scholars appear to differ on whether access to resources facilitates or hinders party-building, their arguments can be reconciled. Patronage and finance have a double-edged effect on party-building: access to these resources weakens incentives but increases the capacity for party-building. Thus, it is relatively uncommon for politicians in power to invest in new parties, but when they do invest in new parties, 70 the state resources at their disposal can provide the means. Moreover, although an initial period in the opposition is usually necessary in order for office-seekers to invest in parties (about which more later), once a party become institutionalized and wins major positions of power, state patronage and public finance usually facilitate organizational expansion (Levitsky, Loxton, and Van Dyck n.d.). 71 Most institutionalized parties initially develop with little access to patronage and finance. Indeed, every new left survivor in Latin America save Chile s PPD spent its formative years in the opposition, with scarce access to state and financial resources. If patronage and finance do not fuel and sustain new parties, what does? Ideology and party-building In an important new study, Stephen Hanson posits that strong ideologies are necessary to sustain collective action and produce durable parties: no ideologies, no parties (Hanson 2010: xv). Hanson argues that French Republicanism, Soviet Communism, and German National Socialism all facilitated party-building by providing politicians and activists with forces for his electoral ground campaigns and tapping the National Intelligence Service for funds and candidates. 70 Politicians in power often, or even typically, invest in new parties for non-electoral purposes like mass mobilization or defense from a perceived threat. See Huntington (1968); Shefter (1994); Smith (2005); Roberts (2006). 71 There are multiple examples of parties that initially developed in the opposition, built grassroots organizations, and, after winning major positions of power decades into their existence, used the state funds and administrative resources at their disposal to continue investing in territorial organization (e.g., Brazil s PT in the 2000s). See Van Dyck (2014); Van Dyck and Montero (n.d.). 18

27 a single, higher cause to coordinate around. These ideologies, on Hansons account, gave members long-term visions of the political future (xxvi) that trumped their short-term electoral or material ambitions and hence artificially [elongated] their time horizons (30). In all three cases, party elites and activists committed to electorally uncertain, poorly resourced party-building projects and stuck it out through thick and thin. Hanson uses the same framework to explain party-building failure in post-soviet Russia. Largely due to the decline of communism, he argues, contemporary Russian parties lack strong ideologies and, consequently, have not managed to sustain elite and activist commitment. While Hanson (2010) makes an important contribution by showing that ideological convictions can extend members time horizons and fuel activist commitment, 72 his argument no ideologies, no parties cannot explain most (or perhaps any) cases of survival among Latin America s new left parties. Although these surviving parties were clearly located on the left of the ideological spectrum, they did not have party ideologies à la the Communists in Russia or Nazis in Germany. On the contrary, given their need to build bridges across different left movements and organizations, and given the global backdrop of communism s decline and collapse, most of these parties, initially, were left fronts. The FSLN, FMLN, FA, PT, and 72 Hanson (2010) states that ideology allow(s) partisans to sustain collective action in the initial phases of party-building (xxii). Ideological convictions, on Hanson s account, lengthen party members time horizons by providing them with clear and consistent visions of the political future. These visions can artificially elongate the time horizons of those who join their cause, making it rational for them to forgo their short-term individual interests in favor of pursuing the long-term benefits available to early converts in the event of an ideological movement s ultimate victory (ibid., xi). Especially under chaotic social conditions, ideologues will usually be the only political entrepreneurs capable of mobilizing large-scale networks of committed activists. This is because political pragmatists, whose political positions...shift quickly along with the rapidly changing political circumstances, will be especially unlikely to sustain long-term commitment to any particular political strategy (ix and p. 79). Thus, ideological parties in turbulent new democracies tend to emerge as the winners in a process of social selection that eliminates their nonideological competitors, which are relatively more vulnerable to the free-rider problem (ibid.: 62). To support his theory, Hanson attempts to show that in Third Republic France, Weimar Germany, and post-soviet Russia all contexts of relative social chaos parties with stronger ideologies (e.g., France s republican party, Germany s National Socialist Party, Russia s Communists) were more likely than pragmatic parties (e.g., France s Orleanists and Bonapartists, German liberal parties, and a multitude of new Russian parties) to build networks of committed activists. 19

28 PRD all brought together a range of conflicting left ideologies spanning orthodox Marxism and pro-capitalist social democracy. Each contained major divisions between moderates and radicals throughout the formative phase. It was not a shared ideology that fueled collective action in these parties. 73 In sum, existing theories of party-building offer valuable insights but cannot explain new left party survival and collapse in Latin America. The remainder of the chapter will present the book s theory of new party survival and failure and set the stage for the coming empirical chapters. The fragile formative phase The formative phase is a critical period, or a hump that new parties must get over in order to achieve institutionalization. 74 In Latin America, the vast majority of parties that emerged after the third wave s onset failed within a decade of their creation. Among the parties that survived the formative years, however, nearly all have become institutionalized, surviving to the present day. On the new left, all six major parties that survived the first fifteen years have survived until the present and remain serious national contenders, 75 some lasting for 73 Hanson identifies the discrediting of Marxism-Leninism as an impediment to the creation of successful parties in post-soviet Russia (xv), but a similar discrediting of Marxism and the resulting elimination of the perceived communist threat did not prevent left (Brazil s PT, Mexico s PRD) and anti-left (e.g., Chile s UDI, El Salvador s ARENA) parties from taking root in Latin America and other third-wave regions. 74 One might call the formative phase a critical juncture, or a relatively short period during which varying conditions across one s units of observations have important long-term consequences. Critical junctures vary in duration, and the key antecedent conditions during critical junctures range from contingent choices by powerful individuals to structural conditions over which individuals have little control (Collier and Collier 1991: 27). The formative phase of party development is relatively long in duration (five to fifteen years), and the formative-phase conditions for new party survival and failure, elaborated above, are primarily structural (i.e., polarization, low access to state resources and mass media, high access to preexisting organizations). 75 Bolivia s MAS, Brazil s PT, Chile s PPD, El Salvador s FMLN, Mexico s PRD, Nicaragua s FSLN. 20

29 over thirty years. 76 In short, the question why only some parties become institutionalized largely reduces to the question why only some parties survive the formative years. Almost without exception, party institutionalization requires the establishment of a successful party brand, or label. Many institutionalized political parties owe their electoral success, cohesion, and durability, almost completely, to their brand. Very few parties, however, are born with a successful brand. While a small subset of new parties do start with strong national brands and electoral bases due to their prior organizational histories, 77 most new parties must first develop their brands by differentiating themselves 78 from other national parties and politicians and, crucially, demonstrating consistency over some period of time. 79 It follows that during the formative phase, most political parties are still in the process of growing and consolidating their brands. 80 As new parties seek to build up and nail down a partisan constituency in the electorate, they almost invariably face electoral disappointments and even crises. New party survival usually depends, thus, on the capacity to overcome these challenges. Grzymala-Busse (2011) has distinguished between regime duration, or longevity, and regime durability, or the capacity to withstand crisis. The book similarly distinguishes between new party survival (i.e., duration, or longevity) and new party durability. New parties tend to fail not because they face crisis, but because they have a low capacity to withstand the setbacks that new parties, survivors 76 FMLN, PT. 77 These include authoritarian successor parties such as El Salvador s ARENA and Chile s UDI (Loxton n.d.) and insurgent successor parties such as El Salvador s FMLN and Nicaragua s FSLN (Holland n.d.). See also Levitsky, Loxton, and Van Dyck (n.d.). 78 They can differentiate themselves through distinctive policy proposals (Lupu forthcoming), sociocultural associations (Ostiguy 2009), mass party structures (Samuels and Zucco n.d.), and more. 79 The differentiation/consistency framework comes from Lupu (forthcoming). 80 Given the importance of differentiation, it is not a coincidence that almost every successful new party in Latin America began with a solidly left (Brazil s PT, Uruguay s FA, El Salvador s FMLN, Mexico s PRD) or right (Chile s UDI, El Salvador s ARENA) program. For a lucid theoretical account of differentiation, consistency, and brand development, see Lupu (forthcoming). 21

30 and failures alike, often face. The theory presented below seeks to explain, fundamentally, why some new parties but not others are equipped to survive crisis. The paradox of adversity: a theory of new party survival and collapse Under what conditions do new parties survive? One might suppose that new parties with national ambitions would benefit from access to state resources and broadcast media. Equally, one might think that new parties would fare best in political atmospheres free of violent conflict and destabilizing mass mobilizations. This book presents the opposite argument. Favorable conditions do not give rise to durable parties. On the contrary, adverse conditions are critical for party-building. Most new parties collapse because they do not have strong organizations with committed activists and, consequently, are ill-equipped to rebound from early electoral crisis. Paradoxically, new parties with strong organizations and committed activists are most likely to emerge under conditions of adversity. Conditions for organizational strength Organizational strength denotes a mass membership and extensive local infrastructure of party branches and nuclei. 81 In a few historical cases, party organizations have encapsulated their bases, structuring members daily lives and personal identities through the sponsorship and organization of everyday activities and the development of distinctive party subcultures (e.g., Europe s early labor and confessional parties). 82 In the contemporary era, strong parties tend to build fewer base-level offices, recruit fewer activists, and play a smaller role in 81 Some have defined organizational strength as the routinization of rules and procedures within the party organization (Levitsky 2003, Krasner 1988, North 1990, Knight 1992), or the extent to which the party organization s members accept the formal and informal rules governing internal decision-making. Others have defined the concept in terms of value infusion, or the degree to which party members value the perpetuation of the party for its own sake, not as a vehicle for the achievement of some separate goal (e.g., economic equality, a particular candidate s election) (Levitsky 2003, Huntington 1968). 82 The mass labor and confessional parties of turn-of-the-century Europe structured millions of members daily 22

31 structuring members day-to-day lives. Still, they continue to have substantial local infrastructures, large activist networks and memberships, and even medium-sized core electorates (Levitsky 2003: 13). Strong organizations deliver votes through several potential mechanisms. First, they facilitate large-scale, on-the-ground electoral mobilization. Party foot soldiers can do campaign work, organizing rallies, going door to door, distributing written information, and transporting individuals to polling booths, while local party offices can provide financial, material and logistical support for these campaign activities. Second, grassroots organizations and activists on the ground can help parties disseminate their brand and develop long-term partisans. 83 Third, territorial organization facilitates the capture of subnational office, which, by allowing parties to gain experience in (and demonstrate a capacity for) government, can contribute to their longer-term success. 84 Fourth, party elites can channel patronage resources more efficiently through party activists (Levitsky 2003; Zarazaga n.d.). 85 Building a strong party organization is laborious, time-consuming, and for party leaders, autonomy-reducing. To begin, top party leaders and local organizers must recruit and formally incorporate large numbers of members and activists. Then, top elites, local leaders, and local activists must house, equip, and staff local offices, establish vertical and horizontal lines of communication across party organs, institutionalize formal or informal mechanisms lives and personal identities, creating entire party subcultures and sponsoring and organizing members everyday activities (e.g., youth leagues, women s groups, sporting clubs, picnics, etc.). See Wellhofer 1979; Sartori 1968; Farneti 1973; Katz 1990; Kalyvas See Samuels and Zucco (n.d.) 84 See Holland (n.d.). 85 Party organizations, in some cases, may provide electorally valuable legitimacy benefits as well (Scarrow 1996: 42). A party s membership statistics may be widely disseminated via mass media; a large membership may act as an effective symbolic representation of a party s existing or target constituencies; a vibrant internal life may enhance perceptions in the wider electorate that the party in question is broad-based, participatory, and internally democratic; in all these ways, a strong organization can make a party more appealing to undecided voters, thus strengthening its electoral performance (Scarrow 1996: 42). (The content in this footnote is repeated, with minor adjustments, in Van Dyck (2014). 23

32 for collective decision-making and conflict resolution, and secure financing for physical infrastructure, communication systems, transportation, and staff salaries, often through membership dues and small donations. These processes take time and, at all levels, require large amounts of volunteer work and the donation and pooling of individual and group resources. As Kalyvas (1996) observes, organization-building does not come naturally or automatically to political actors. It is a difficult, time-consuming, costly, and often risky enterprise (41). Crucially, electoral progress based on organization-building occurs slowly. Large activist bases diminish elite nimbleness and autonomy, imposing procedural constraints and demanding ideological stances that inhibit rapid electoral growth (Kitschelt 1994: 225). Office-seekers only have an electoral incentive to invest in organization if they cannot win office via lower-cost routes, or routes that require less sacrifice in time, labor, and autonomy. Two such routes are (1) the distribution and mobilization of state resources and institutions (Hale 2006) and (2) mass media appeals. Consequently, the adverse and relatively rare combination of low access to the state and low access to mass media creates an electoral incentive for office-seekers to invest in party organization. State, media, and incentives for party organization As noted earlier, in reference to Hale (2006), politicians in power have incentives to use the state for electoral purposes rather than undertaking the costly, autonomy-reducing work of party-building. To date, however, few scholars have focused on the positive implications of this insight for party-building. Simply put, because access to state resources weakens incentives for party-building, low access to state resources strengthens incentives for partybuilding, other things equal. For office-seekers, an extended period in the opposition actually serves as a blessing in disguise. In the political wilderness, individuals cannot tap state coffers and institutions for candidates and campaigners. Effectively, they lack access to a major 24

33 party substitute one which has prevented politicians from building parties in numerous third-wave countries. In the contemporary era, office-seekers without access to the state often do not need baselevel organization in order to obtain a mass following. 86 Given the broad penetration of television and radio in most countries today, media-savvy political entrepreneurs can win major elections, including presidential elections, through mass media appeals, with little or no party organization behind them. 87 Moreover, they can do so almost instantaneously, with great autonomy, and with minimal effort or resources. 88 The rise of broadcast media has reduced elite incentives 89 for party-building and, consequently, weakened or prevented the emergence of strong parties in many present-day polities. 90 Mainwaring and Zoco (2007) find that contemporary party systems are more volatile 86 In addition, in the contemporary period, high access to the state typically brings regular and extensive broadcast media coverage. Thus, state-originated parties and candidates often use broadcast media to maintain and build electoral support. 87 Hale (2006) makes this observation succinctly: In the era of mass communications and widespread media, it is entirely possible for a party to win office almost solely on the basis of ideational capital (i.e., without patronage or organization) (14). 88 Levitsky and Cameron (2003) state that contemporary politicians may reach millions of voters through television and may do so more quickly and at lower cost (in terms of human and organizational resources) than through party organizations (24). 89 Mass media also change ordinary citizens incentives. In the age of widespread media use, party membership has become less essential for individuals seeking political information or channels of communication with elites. Further, there is evidence that broadcast media consumption causes individuals to disengage from civil society (e.g., local party organizations) (Putnam 1995: ). Katz (1990) finds evidence of these effects, identifying a strong inverse relationship between individuals TV/radio consumption and party membership (157). 90 This argument should not be overstated. The growth of mass media has not entirely eliminated electoral incentives for organization-building (Van Dyck n.d.). Party organization continues to give parties an important advantage in local elections, where mass media campaigns tend to be less cost-effective and more logistically difficult: (t)elevision debates and advertisements are more easily planned (and, in the case of paid advertising, cheaper to produce) the fewer the locally based interests to which special appeals have to be made (Ware 1992: 74). Moreover, in larger elections, vote-seeking elites with access to mass media have at least some incentive to pursue the additional votes (i.e., the marginal electoral benefits) that result from separate strategies, like organization-building (Scarrow 1996: 36; Hale 2006: 206; Epstein 1980: 375). Still, for individuals with access to media, the marginal electoral benefits of organization-building are typically insufficient to justify the considerable associated costs in time and labor. 25

34 than historical party systems due, in part, to the emergence of television as a major actor in enabling candidates to win election... It is easier and in the short term more effective to use the modern mass media than to build a party (156-7). 91 Katz (1990) observes that in contemporary Western Europe, the party meeting, the party canvasser, the party press [are] all supplanted in importance by the party leader speaking directly to his or her supporters on the small screen (146). 92 Since the collapse of the Soviet Union, media-based, personalistic political vehicles have proliferated in Russia, 93 but not a single strong party has taken root. 94 In South America during the 1990s, three new left parties Colombia s M-19, Argentina s FREPASO, Chile s PPD rose to national prominence through mass media appeals, with weak or nonexistent organizations. 95 Fernando Henrique Cardoso, leader of Brazil s PSDB, memorably observed in the late 1980s that a TV channel is worth more than a party (Mainwaring 1999: 150). 96 In short, for individual office-seekers, state- and media-based electoral strategies, relative to organization-based strategies, exact minimal costs in time, labor, and autonomy. These strategies allow for elite nimbleness and autonomy and bear electoral fruit quickly. Consequently, politicians and political entrepreneurs with access to the state or media tend not to invest in mass party structures, instead using state resources and institutions as party 91 Mainwaring and Zoco (2007) also identify the earlier parties incorporation of large blocs of new citizens as a key factor. In this connection, see also Coppedge Landi (1995) similarly states that...television, radio, and the media in general outstrip the intermediating function of the local party organization. [The party activist] has become dispensable (211-2). 93 E.g., the Liberal Democratic Party of Russia, Fatherland-All Russia. 94 Hale (2006): 66-8, Hale (2006) writes of post-soviet Russia that with television sets in every home...a political entrepreneur no longer needs a mass organization of the kind traditionally wielded by parties to reach large numbers of voters. Instead, one simply needs enough money to purchase advertising time or a flair for obtaining coverage on television news (242). 95 Boudon (2001); Abal Medina (1998, 2009); Plumb (1998). 96 Fernando Collor de Mello, the telepopulist, won Brazil s 1989 presidential election largely due to televised advertisements, appeals, and debate performances. His flash party, the PRN, had no grassroots structures, although he did benefit from the support of allied right-wing mayors with control over local patronage (Ames 1994). 26

35 substitutes or attempting to build electoral constituencies through mass media appeals. Crucially, politicians and office-seekers who use state resources or mass media to build support tend to rise relatively quickly. As they gain visibility and support, they close their share of the electoral market, eclipsing or crowding out programmatically similar, would-be organization-builders. Ideologically proximate elites and activists who might otherwise have preferred a strategy of grassroots territorial expansion must either jump on the bandwagon of the ascendant state-/media-based party or remain electorally marginal. 97 In these contexts, new grassroots parties do not have the time and electoral space necessary to develop and grow. Yet where large segments of the opposition lack access to media, the only new opposition parties capable of achieving electoral liftoff are those that build strong organizations. Before the mass media age, externally created parties could not challenge established regimes or governing parties except by building strong party organizations (Shefter 1994). 98 Historical examples of mass-based opposition parties thus abound and span the ideological spectrum. Classic examples include the mass confessional parties of Belgium, the Netherlands, Austria, Germany, and Italy, which formed in opposition to Liberal regimes espousing anticlericalism, 99 and their labor-based counterparts, which formed in opposition to regimes and governing parties perceived as hostile or indifferent to organized labor (e.g., the British Labour Party, the German Social Democratic Party). 97 Argentina s FREPASO provides a case in point. In the 1990s, Argentina s FREPASO, a new left party with a weak organization, experienced a meteoric rise, driven almost entirely by several leaders especially Carlos Chacho Álvarez s mass media appeal (Palermo and Novaro 1998; Abal Medina 1998, 2009). Many Argentine left-wing activists supported the fast-growing, media-based FREPASO even though, in interviews, they uniformly expressed a preference for organizational strategies over media-based ones. One interviewee noted that the only alternatives to FREPASO were tiny, electorally marginal left parties (See Chapter Six). 98 Shefter argues that parties founded by regime outsiders (externally mobilized parties) build strong territorial organizations based on ideological, programmatic appeals. 99 Kalyvas (1996). 27

36 In the contemporary era, many opposition parties enjoy access to broadcast media, but not all do. In some of the world s least developed countries, broadcast media reach too few people to form the basis of a national electoral strategy. 100 More commonly, authoritarian restrictions on freedom of the press foreclose the media option for contemporary opposition parties. These parties must invest in territorial organization in order to contend for national power. Mobilizing structures and the capacity for organization Incentives are necessary but not sufficient for organizational strength. Party-builders with incentives for organization must have the capacity, or means, 101 to recruit masses of members and implant party infrastructure in large swathes of territory. The means for organizationbuilding come, primarily, 102 from external mobilizing structures, especially preexisting organizations and movements with territorial reach, large memberships, ready-made organizational hierarchies, cadres of experienced organization-builders, and visibility and legitimacy among segments of the wider electorate. 103 Without access to such mobilizing structures, office-seekers will find it difficult or impossible to build strong party organizations. 104 Although states sometimes serve as platforms for strong party organizations (cf. Shefter 1994), 105 mobilizing structures for party-building typically come from civil society. His- 100 CITATION NEEDED. 101 Levitsky, Loxton, and Van Dyck (n.d.). 102 Levitsky, Loxton, and Van Dyck (n.d.) also identify finance as a useful means. 103 Kalyvas (1996); Kitschelt et al. (1999); LeBas (2011); Vergara (2011); Ziblatt (n.d.). Party-builders work[] with the materials at hand and buil[d] the organizations that [are] possible (LeBas 2011: 52). 104 Preexisting organizational infrastructure also facilitates state-building. In his comparison of Italian and German unification, Ziblatt (2006) argues that Germany, unlike Italy, was able to build a strong federal state because preexisting subnational governments possessed infrastructural power. Infrastructural power at the subnational level made it feasible for the central German state to devolve state functions to subnational units. Italy created a unitary state due to subnational infrastructural weakness. 105 See also Levitsky, Loxton, and Van Dyck (n.d.); Loxton (n.d.). 28

37 torically, mass parties have grown out of civil society actors such as trade unions, social movements, grassroots church networks, and even guerrilla armies. These feeder structures donate vital assets, including base-level activist networks, experienced elites, institutionalized chains of command, physical locales, office and communications equipment (e.g., computers, telephones, fax machines), and vehicles. Mobilizing structures Group builds organization Group seeks national office No media, no state Electoral success requires organization No mobilizing structures Group does not build organization Media or state Electoral success does not require organization Group does not build organization Figure 1.1: Conditions for organizational strength A number of scholars thus argue that party-building outcomes largely depend on the strength of available civil society feeder organizations. Kalyvas (1996) argues, for example, that territorially extensive, mass-based, lay church organizations provided Christian Democratic party-builders in Europe with electorally valuable activist networks and vital infrastructure. 29

38 Where these church networks existed, strong confessional parties formed (e.g., Germany). 106 Where they did not, strong confessional parties did not emerge (e.g., France). Along similar lines, LeBas (2011) identifies the presence of autonomous labor unions as the key factor determining whether parties with organizational reach emerged in democratizing West Africa. 107 Others have attributed divergent party-building outcomes to crossnational variation in social density generally, 108 or to the differential strength of specific feeder organizations such as guerrilla armies, 109 ecological movements, 110 and even business conglomerates. 111 Polarization and the sources of activist commitment Organizational strength is necessary but not sufficient to equip new parties for survival. To be durable, new parties, even those with territorial infrastructure and large activist networks, depend on high levels of activist commitment. In recent work, Steven Levitsky has argued that settings of intense societal polarization and conflict facilitate the emergence of committed party activists, or partisan believers (Panebianco 1988: 26-30), on a national scale. 112 Conditions of low state and media access select for, but do not generate, committed activists. Parties born in the state attract patronage-seekers. Since media-based new parties can 106 Also, Belgium, Netherlands, Austria, Italy, and other countries. 107 Such associations provided organizational resources, independent ties to mass constituencies, and a great deal of political visibility, constituting mobilizing structures that could be co-opted by opposition parties, giving these parties a greater ability to mobilize voters across the lines of ethnicity and geographic space (LeBas 2011: 51). 108 See Vergara (2011) on party development in Bolivia (high social density, relatively strong parties) and Peru (low social density, relatively weak parties). 109 Nicaragua, El Salvador (strong) v. Honduras, Guatemala (weak) (Allison 2006). 110 West Germany (strong) v. Belgium (weak) (Kitschelt 1989). 111 See Barndt (n.d.) on Panama. 112 Levitsky, Loxton, and Van Dyck (n.d.). 30

39 reach millions instantaneously and achieve meteoric electoral rises, they too tend to attract careerists. In addition, they often have an incentive to appeal to the median voter rather than a narrow, radical base. 113 For these reasons, both state-originated parties and media-based new parties typically claim relatively few hardcore, ideologically driven supporters. Parties born with low access to state and media do not, by and large, attract patronageseekers and short-term opportunists. Material resources are usually limited, 114 and electoral progress is uncertain and slow. 115 Consequently, party elites cannot offer selective incentives (e.g., jobs, salaries) to grassroots workers in the short term. 116 Low state and media access thus weeds out opportunists seeking short-term patronage. As a rule, only ideologically committed activists, or what Panebianco (1988) calls believers (26-30), are willing to volunteer their time and labor for a new party with scarce financial resources and weak, uncertain electoral prospects (also see Shefter 1994; Greene 2007). 117 Yet selection pressures do not generate activists. There must be a positive goal, or cause, for party supporters to believe in and rally around. In Panebianco s terms, there must be a 113 Argentina s FREPASO provides a case in point. See Chapter Six. 114 Exceptions include new conservative parties with strong ties to the economic elite (Loxton n.d.; Luna n.d.). 115 These parties do not become serious contenders for national power quickly. Only by developing a mass membership, large activist base, and territorial infrastructure, and by bringing base-level demands into harmony with broad-based appeals, can they compete at the national level. 116 In successful, established parties, leaders often motivate party workers by providing selective incentives, especially government and party jobs. 117 Shefter (1994) argues that where elites seek to mobilize a mass party and lack access to state resources, they must mobilize individuals on the basis of ideological appeals. Greene (2007) posits that in dominant party systems, early joiners of opposition parties must be ideologues, as only ideologues are willing to engage in party work for long periods despite the low short-term probability of unseating the dominant party. More specifically, in dominant party systems, new parties face a massive resource disadvantage, which greatly reduces the likelihood of short-term opposition success and thus prevents patronage-seeking individuals from committing to the opposition. [D]eep anti-status quo beliefs, he holds, are a near necessary condition for the survival of opposition parties in dominant party systems (6). He shows that the elites and activists who built Mexico s PAN and PRD the country s two main opposition parties under PRI hegemony had comparatively extreme policy preferences and ideological commitments. 31

40 collective incentive that leaders can furnish, in lieu of patronage, to motivate activists (9). 118 As noted earlier, Hanson (2010) treats ideology as a potent collective incentive, necessary for sustaining member commitment, and hence essential for building strong parties. It is rarely possible, however, to build a territorially extensive, activist-based party organization around a single ideology. In order to build such an organization, new party leaders, in the typical case, must make room for a large number of local leaders, networks, movements, and organizations with distinct ideological, regional, class, and even cultural profiles. 119 Conditions of intense societal polarization and conflict, according to Steven Levitsky, are more useful for producing large-scale partisan activism. 120 Periods of populist mobilization and counter-mobilization, movements against authoritarian regimes, and life-or-death struggles such as revolution and civil war can unite otherwise disconnected groups around higher causes and facilitate collective action. These social struggles spur heterogeneous actors to join forces, leading to the crystallization of broad, new political identities. In this way, the polarization associated with contexts of struggle can generate partisan believers, or militantes, on a national scale (Levitsky, Loxton, and Van Dyck n.d.). Polarization and conflict, when violent, also intensify selection pressures (Levitsky, Loxton, and Van Dyck n.d.). Risking one s life or well-being in the service of a political cause requires extraordinary commitment and belief. Individuals with weak ideological commitments are loath to participate in revolutionary struggles, civil wars, or movements against highly repressive authoritarian regimes. New parties born in these contexts tend to attract individuals whose convictions trump their risk aversion and desire for short-term political or 118 See also Shefter (1994); Greene (2007); Hanson (2010). 119 LeBas (2011) notes, for example, that in most contexts, pro-democracy movements try to incorporate all actors outside the state: inclusion...creates strong tendencies toward fragmentation. For the opposition parties that succeed movements, the challenge is...[to prevent] the defection of elites and constituencies (LeBas 2011: 58). 120 Levitsky, Loxton, and Van Dyck (n.d.). 32

41 material gain. Like resource scarcity and electoral uncertainty, violent conflict thus selects for committed activists. Surviving electoral crisis So far, the section has argued that four conditions low access to the state, low access to media, access to mobilizing structures, and high levels of polarization and conflict facilitate the emergence of new parties with strong organizations and high levels of activist commitment. To recap, given the relatively high costs of organization-building in time, labor, and leader autonomy, office-seekers only have an electoral incentive to invest in mass party structures if they lack access to state resources and mass media. Conditions of low access to resources and media weed out political opportunists, selecting for committed activists disposed to carry out grassroots party work absent a short-term material payoff. Incentives alone do not produce strong party organizations, and selection pressures alone do not produce committed activists. Access to mobilizing structures such as guerrilla armies, trade unions, and social movements provide the capacity, or means, for organization-building. Situations of intense polarization and conflict, such as civil war or populist mobilization, help generate committed activists by producing the higher causes, or collective incentives, that spur individuals to collective action. Without organizational strength and activist commitment, new parties tend to be fragile. Parties without base-level organization (e.g., many media-based parties) depend for their early survival on the establishment of a successful party brand and, where applicable, strong performance in government. 121 If their brand becomes diluted due to programmatic shifts or the emergence of a strong competitor, or their credibility to govern becomes tarnished due to weak performance, they have no organization to fall back on and are likely to fold (E.g., Argentina s FREPASO, Colombia s M For a similar point, see Cyr (n.d.). 33

42 In parties based on patronage, or on the expectation of short-term access to patronage, party members do not coordinate around a higher cause, but in order to advance their individual careers. Even if these parties possess base-level organization, elites and base-level party workers lack robust commitment. Consequently, individuals within the leadership and rankand-file are more likely to defect if the party performs poorly in elections (Levitsky and Way 2012). 123 New parties with extensive, organized networks of committed activists are more durable than their counterparts, capable of rebounding from the electoral disappointments and crises that commonly beset incipient political projects. Driven by a cause rather than a material payoff, party activists in territorial strongholds have long time horizons, which are critical amid electoral crisis. If these parties have a terrible election or spend a long period in the wilderness, activists are more likely to stick it out. Defections are less frequent and severe, and if major defections do occur, outright collapse is less likely to occur, as a core of members will typically remain active, and hardcore voters will typically remain loyal. 124 In short, although extensive networks of committed activists may impede vote maximization, 125 they make new parties durable, giving them something to fall back on amid early crisis. Moreover, since new parties with strong organizations claim local activists and often cannot contest national power in the early years, they tend to prioritize subnational government, seeking to elect mayors, municipal councilors, state and federal legislators, and even state or provincial executives (e.g., governors) in their local and regional bastions. Subnational gover- 123 Parties that are organized exclusively around patronage and career ambition may effectively discourage defection during normal times, while the party s hold on power is perceived as secure. However, such parties are vulnerable to crisis, or any exogenous shock that threatens their capacity to deliver the goods (Levitsky and Way 2012: 8). 124 In the extreme forms mentioned earlier (e.g., early twentieth-century labor and confessional parties in Europe), mass organization also greatly reduces the frequency and intensity of electoral setbacks; the more individuals a party encapsulates, the higher and more stable a party s electoral floor will be. 125 See Greene (2007), (n.d.). 34

43 nance allows for greater policy flexibility and experimentation 126 and also, given its relatively small scale, poses less of a management challenge than national governance. Consequently, winning subnational elections gives new parties a valuable opportunity to establish distinctive brands, develop strong performance records, and thus consolidate electoral constituencies at the subnational level (Holland n.d.). New parties with solid local constituencies often retain important subnational executive posts despite national-level electoral disappointments and setbacks. Thus, in the event of national crisis, new parties that have invested in subnational elections and governance are more likely to endure, other things equal. 127 Conditions for avoiding fragmentation and schism Although large networks of committed activists fortify parties during electoral crisis, they also increase the risk of schism. 128 Ideological political movements are prone to fragmentation. When factions prioritize doctrinal purity over political pragmatism and compromise, the risk of debilitating schism is high. Consequently, many ideological parties are plagued by sectarianism and internal divisions (e.g., Latin American communist parties). Moreover, it is rarely feasible to build a strong organization without incorporating different ideologies, regions, and classes. For these reasons, avoiding fragmentation and schism is a central chal- 126 Brazil s PT, for example, used innovative programs like participatory budgeting to consolidate subnational constituencies in different parts of the country (e.g., Porto Alegre). 127 It should also be noted that new party elites, in this situation, can continue providing selective incentives to activists by distributing municipal- or state-level government jobs. The dispensation of lower-level public-sector positions in subnational strongholds may, in some cases, play a minor role in new parties survival of early electoral crisis. 128 There are many ways for new parties to collapse. Most cases of collapse, however, fall into one of two broad categories: electoral collapse and schism. In cases of electoral collapse, new parties experience prolonged electoral disappointment or major electoral setbacks and do not recover. In cases of schism, internal conflicts arise, major players defect, and new parties disintegrate as a result. Though analytically separate, electoral failure and schism are often causally related. If a new party suffers electorally, party elites and factions become more likely to split off. Conversely, if a new party loses major elites or factions, it becomes more likely to suffer electorally. 35

44 lenge for mass-based parties (LeBas 2011). 129 What factors prevent debilitating schisms in new parties, especially those with large, committed activist networks? Violent conflict as a source of cohesion In addition to selecting for and generating committed activists, violent conflict, according to Levitsky and Way (2012) and Lebas (2011), can decrease the likelihood of fragmentation and schism within participating movements and parties. On this account, violence between groups creates dynamics within groups that raise the cost of defection. Such violence sharpens we/them distinctions and creates perceptions of linked fate among competing cadres (Levitsky and Way 2012: 871). Defection may come to be viewed as an act of treason at the base level. Group leaders may reinforce social incentives against defection by labeling group moderates as traitors. Where the groups in question are political parties, voters may even punish elite splinter groups at the ballot box, effectively trapping potential defectors within the party (LeBas 2011: 46). 130 To support this argument, both Levitsky and Way (2012) and LeBas (2011) provide evidence from sub-saharan Africa. Party systems, they show, solidified where regime and opposition engaged in violent conflict (Zimbabwe, Mozambique) but fragmented where such violence did not occur (Kenya, Zambia) In her recent analysis of party strength and weakness in democratizing East Africa, LeBas (2011) emphasizes that [i]t is not enough for parties to win mass support... (T)hey must develop means of resolving conflict within their organizations (and) preventing party fragmentation. (35). Later, she writes that pro-democracy movements try to incorporate all actors outside the state: inclusion...creates strong tendencies toward fragmentation. For the opposition parties that succeed movements, the challenge is...[to prevent] the defection of elites and constituencies (LeBas 2011: 58). 130 LeBas (2011) generalizes that in highly polarized societies, large blocs of the electorate typically identify with one of the conflicting groups. As a result, politicians who run independently of parties representing the poles, or who defect from such parties, are unlikely to achieve electoral success. Polarization thus creates electoral pressure for elites not to defect, even amid internal conflict. She writes: Where party competition is polarized, independent and third-party candidates are significantly less likely to win election battles, and political aspirants are therefore unlikely to form new parties, even if they lose nomination battles. Polarization creates intraparty cohesion by trapping potential defectors within the party (46). 131 LeBas (2011) argues that despite numerous structural and institutional similarities between Zimbabwe and 36

45 Although Levitsky and Way (2012) and LeBas (2011) demonstrate that life-or-death struggles can generate cohesion in new parties, many new parties do not develop under conditions of intense, violent conflict but still manage to cohere. For example, of Latin America s six new left survivors, only two, during or prior to formation, acted as belligerents in systematically violent national conflicts (FSLN, FMLN), and arguably, these two parties depended primarily on their strong electoral brands which they were born with 132 to avert fatal schisms. The other new left survivors did not begin with strong brands and either did not suffer violence or experienced irregular, localized violence that left the core leadership unharmed. What factors might have prevented fragmentation and schism in these parties? Indispensable leadership: coattails and unquestioned internal authority Theorists of party-building rarely take the role of individual leaders seriously, perhaps to avoid perceptions of excessive voluntarism. Yet leaders often have a decisive impact on new parties fortunes, particularly in presidential systems (Levitsky, Loxton, and Van Dyck n.d.). Leaders play an especially important role in reducing the likelihood of debilitating internal conflicts and fatal schisms. 133 New parties typically lack two features that generate cohesion in established parties: (1) a strong party brand, and (2) institutionalized mechanisms for resolving internal conflicts. Almost all established parties possess a strong brand, which creates a powerful electoral Kenya, only in Zimbabwe did the regime engage in violent conflict with the opposition. The conflict gave rise to intense polarization, empt[ying] out the middle ground between the incumbent party (ZANU- PF) and the primary challenger (MDC). In their study of ruling party elite cohesion, Levitsky and Way (2012) reach a similar conclusion, finding that in Zimbabwe and Mozambique, but not in Kenya or Zambia, collective experiences of violent struggle proved a critical source of cohesion for Zimbabwe s ruling ZANU-PF and Mozambique s FRELIMO. 132 Levitsky, Loxton, and Van Dyck (n.d.). 133 Popular leaders also help new parties achieve electoral viability. See Levitsky, Loxton, and Van Dyck (n.d.). 37

46 incentive against defection. Elites in established parties know that if they defect and run for office independently, they will lose the partisan vote and probably suffer electorally. In addition, established parties often possess institutionalized mechanisms for conflict resolution. With time, and in response to internal pressures, many parties, especially those with large memberships, develop formal channels for competing interests and factions to promote policy positions and candidates, as well as accepted rules for selecting candidates and hammering out the official party program. As noted earlier, a small subset of parties are born with strong brands in virtue of their prior organizational histories (e.g., FMLN). Most new parties, however, do not possess a brand capable of providing strong electoral incentives against defection. Equally, most new parties especially heterogeneous, mass-based new parties lack institutionalized mechanisms for resolving disputes. Horizontal linkages tend to be weak. Constituent factions or parties typically have not worked together and thus lack norms of cooperation and established procedures, formal or informal, for adjudicating internal conflict. A party leader who combines powerful electoral coattails with unquestioned internal authority can generate cohesion in new parties by substituting for the missing brand and conflict resolution mechanism. When a party leader has more external electoral appeal than any other party elite, lower elites know that if they defect and run for office separately, they will be less likely to win, unable to ride on the coattails of the leader. Such a leader typically a candidate for top executive office 134 substitutes for a strong party brand by providing an electoral incentive against defection. History furnishes numerous examples of heterogeneous parties and coalitions whose leaders, in virtue of their external electoral prospects and clout, played crucial roles in preventing the defection of competing elites and factions I.e., the presidency or a major governorship or mayoralty. 135 These include Charles de Gaulle s RPF (est. 1947), a big tent party encompassing socialists, communists, 38

47 Electoral coattails also give the party leader a significant degree of leverage in internal disputes. When the party leader creates a powerful coattails effect, lower elites and opposition factions typically want him to remain in the party, for electoral reasons at least. Thus, during internal struggles and negotiations with the leader, they will be inclined to make more concessions and fewer demands than they might otherwise. Leaders tend to use this leverage, to the extent possible, to shape party policy and behavior as they see fit. Yet coattails alone are a weak source of internal authority. If a new party leader does not possess any additional source of authority, and must lord his external clout over major wings of the party, he will have a limited capacity to impose on the party during critical periods of internal tension and conflict. In order to remain united during these critical periods of internal struggle, new parties that lack established conflict resolution mechanisms require a leader with unquestioned, virtually unassailable authority. Otherwise, internal conflicts will escalate and may devolve into anarchy, with major opposing players reserving veto power given the absence of a universally accepted arbitration mechanism. A leader with unquestioned authority can step into this power vacuum and serve as the party s final arbiter, or informal high court, mediating diverse demands and, when necessary, imposing a contentious decision that all accept. In this way, such leaders enable otherwise divided parties, during critical, delicate moments, to move forward with a minimum of coherence and unity, speaking in a single voice. Where does this kind of internal authority come from? Aside from the internal leverage that comes with coattails, there are two broad sources of internal authority, which are not mutually exclusive or, in some cases, even independent. The first source is strong crossfactional ties, meaning productive, constructive relationships with the party s main elites Christian Democrats, and conservatives; Juan Perón s PJ, which included strong left-wing pro-labor elements, centrists, and conservatives; and new left parties and coalitions such as Brazil s PT under Lula, Peru s IU under Alfonso Barrantes, and Mexico s PRD under Cuauhtémoc Cárdenas. 39

48 and factional leaders. Particularly in mass-based, internally riven new parties, leaders who possess strong crossfactional ties often become indispensable (Ansell and Fish 1999). By establishing positions at the hubs of large, complex party organizations, these leaders develop a singular capacity within their parties to balance conflicting interests and broker difficult deals. 136 Since their parties rely on them for unity, these leaders amass considerable power internally. In numerous parties characterized by major internal divisions and weak horizontal linkages, particular elites have risen to the position of party leader due to their strong crossfactional ties: Helmut Kohl of West Germany s Christian Democratic Union (CDU), 137 Zennady Zyuganov of the Russian Communists, François Mitterand of the French Socialists, 138 Carlos Menem of Argentina s Justicialist Party (PJ), 139 and others. 140 In new parties, the factional ties that the party leader brings to the table at time zero are critical. In some cases, leaders have already played active, leading roles in the party s formative struggles and main feeder organizations. These leaders begin with strong factional 136 If one visualizes a party as a network in which major factions and elites are represented by nodes, leaders with crossfactional ties act as the hub, with spokes connecting them to the major nodes. See Padgett and Ansell (1993); Ansell and Fish (1999); Levitsky (2003). 137 During the 1960s and 1970s, the CDU was riven with internal divisions along several dimensions, including region, religion, and the party s stance regarding East Germany and reunification (Ansell and Fish 1998: 295). According to Ansell and Fish (1998), these divisions generated conflicts that might have led to major schisms if not for the networking of party chairman and eventual chancellor, Helmut Kohl: The basis for Kohl s success lay in his remarkable talent as an integrator and mediator... Even before winning the party chairmanship in 1973, Kohl had staked out a role as equilibrator and as linchpin of party unity... When the party was torn by internal strife, Kohl consistently managed to embody the party itself rather than a tendency or a faction (ibid.: 294). 138 Ansell and Fish (1999). 139 During the 1990s, the PJ was extremely diverse, with elites and factions spanning the far left and far right of the political spectrum. Levitsky argues that party leader Carlos Menem, due to his strong ties across Peronism s complex system of internal networks, became indispensable to Peronist unity (169). Menem, during his political career, undertook a series of alliance shifts that took him from the far left of the party to the far right, and in his pursuit of the 1988 presidential nomination, Menem managed to build a coalition...including old-guard unionists, Orthodox leaders, and rump cadre organizations of the left...and right that few leaders, if any, could match (Levitsky 2003: 170). 140 E.g., Cosimo de Medici of the Medicean Party (Padgett and Ansell (1993). 40

49 ties (e.g., Lula da Silva in Brazil s PT). 141 Other individuals rise to positions of party or coalition leadership precisely because they lack ties to any faction and, consequently, do not threaten to tilt the internal balance of power in a particular faction s direction (e.g., Alfonso Barrantes of Peru s IU). 142 It is generally harder for the latter type of leader to establish strong crossfactional ties during the party s formative years. Internal authority is not solely a function of the leader s crossfactional ties. It is also a function of the leader s moral stature among party members. Leaders vary in the degree of respect, credibility, and moral legitimacy that party members accord them. Some party leaders acquire leadership positions in new parties without moral legitimacy among large segments of the base. This may occur most often when heterogeneous new parties purposefully designate a leader who lacks factional ties of any kind. Other new party leaders enjoy broad respect and credibility within their parties. Among these, some may even possess a mystical quality and be seen by party members as central to the party s identity, or as the embodiment of its higher cause. This kind of stature within a party may be rooted in factors such as symbolic pedigree, 143 acts of heroism, 144 experiences of hardship, 145 leadership in the party s founding struggles, 146 and more. Insofar as a leader enjoys credibility, commands respect, or even inspires devotion within his party, he will find it easier to assume the role of internal arbiter when necessary See Chapter Three. 142 See Chapter Seven. 143 E.g., Cárdenas of Mexico s PRD. 144 E.g., Charles de Gaulle of France s RFP. 145 E.g., Nelson Mandela of South Africa s ANC. 146 E.g., Lula of Brazil s PT; Robert Mugabe of Zimbabwe s ZANU. 147 High moral stature internally may even facilitate the forging of crossfactional linkages. Leaders who seek to build and nurture ties to previously unknown leaders and activists from across the spectrum may be more likely to succeed if they enjoy broad credibility or possess moral legitimacy. Otherwise, the forging of crossfactional linkages may prove more challenging and depend on the individual s capacity and disposition for assiduous networking and coalition-building. Relevant examples include Helmut Kohl 41

50 In order for leaders to attain maximum authority within their parties, they must possess strong crossfactional ties and high moral stature. If a leader combines these qualities with the leverage that accompanies coattails, his capacity to arbitrate or impose during critical internal struggles may prove decisive in overcoming internal divisions. Coattails Internal authority Barrantes Lula, Cárdenas Handal, Zyuganov Figure 1.2: Coattails and internal authority 148 Crucially, there is no necessary relationship between a party leader s external electoral appeal and internal authority (Figure 2). Historically, most party leaders have fallen into one of three categories: those with strong internal authority but relatively low levels of external of West Germany s CDU, Gennady Zyuganov of Russia s Communist Party (Ansell and Fish 1999), and Carlos Menem of Argentina s PJ (Levitsky 2003: ). 148 Alfonso Barrantes of Peru s IU (Chapter Seven); Lula da Silva of Brazil s PT (Chapter Three); Chauhtémoc Cárdenas of Mexico s PRD (Chapter Five); Schafik Handal of El Salvador s FMLN; Gennady Zyuganov of Russia s CPRF. 42

51 appeal, 149 those with high levels of external electoral appeal but weak internal authority, 150 and those who combine high levels of external electoral appeal with unquestioned internal authority. 151 In some new parties, internal authority and the power to provide coattails reside in competing elites: one with the most external electoral clout, the other with the strongest crossfactional ties or highest moral stature. Intraparty tension tends to escalate in these cases due to the existence of two competing power centers. 152 In summary, new parties with strong organizations and committed activists are prone to fragmentation and schism. While violent conflict can create dynamics within these parties that raise the cost of defection, most new parties with strong organizations remain intact in the absence of violent conflict. This section has argued that individual leaders can function as vital sources of cohesion. Party leaders who combine unique external appeal with unquestioned internal authority may help to bind embryonic parties together during the critical formative period. Because their word is effectively law, they can step in during moments of conflict and paralysis and impose a path forward. By furnishing coattails, they discourage defection and reduce the likelihood of debilitating schism. In this way, strong party leaders may substitute for established brands and institutional mechanisms of dispute resolution during parties formative periods. The role of authoritarianism The chapter has identified a set of conditions for new party survival: low access to state resources, low access to media, access to mobilizing structures, polarization and conflict 149 E.g., Álvaro Alsogaray of Argentina s UCeDé, Gennady Zyuganov of Russia s Communist Party, Manuel Fraga of Spain s People s Party, Schafik Handal of El Salvador s FMLN. 150 Alfonso Barrantes, the leader of Peru s United Left (IU), belonged to this category during the IU s formative phase. 151 The empirical chapters will argue that Lula da Silva of Brazil s PT and Cuauhtémoc Cárdenas of Mexico s PRD both belonged to this category during the PT s and PRD s formative phases. 152 This scenario obtained in Peru s IU during the 1980s. See Chapter Seven. 43

52 (violent or nonviolent), and a party leader who combines coattails with unquestioned internal authority. None of these conditions is necessary for the formation of durable parties, and jointly, they are not necessarily sufficient. Each, however, increases the likelihood of partybuilding success, other things equal. 153 Three of the above conditions low access to state resources, low access to mass media, and polarization/conflict are loosely associated, empirically, with opposition party-building under authoritarian rule. First, new parties under authoritarian rule are more likely to develop with minimal access to state resources. In most authoritarian systems, 154 opposition parties can operate. In these systems, regimes use different tactics to exclude the opposition from state structures. Some simply ban elections. During the era of bureaucratic authoritarianism (BA) in Latin America, for example, right-wing military regimes in the Southern Cone eliminated elections at all levels of government. 155 Uruguay s Broad Front (FA), for example, spent eleven of its first thirteen years without contesting a single election or occupying a single public office. In other cases, regimes allow elections but tilt the electoral playing field against opponents. These competitive authoritarian regimes use fraud, repression, exile, surveillance, financial advantage, and judicial tampering to weaken the electoral opposition and retain a monopoly on state resources (Levitsky and Way 2010). Second, authoritarian regimes handicap opponents by limiting their access to media. Through various mechanisms, including selective licensing, bribes, sanctions, and outright ownership, 153 Specifically: New parties that lack access to state resources and mass media are more likely than their counterparts to invest in organization. If these same parties originate in struggle, they are more likely to have activist commitment, due to high polarization levels, and the means for organization-building, through access to mobilizing structures (e.g., trade unions, guerrilla armies). New parties with strong organizations and high levels of committed activists are more likely to weather electoral crisis. Finally, new parties are less likely to fragment and break apart if the party leader combines coattails with unquestioned internal authority. They are also less likely to split apart if they emerge from violent conflict. 154 Extremely closed regimes (e.g., North Korea, Cuba) make it impossible for opponents to organize. 155 (Elections were banned in Argentina during the late 1960s and early 1990s, in Chile during the 1970s and 1980s, and in Uruguay during the 1970s and early 1980s. The Brazilian BA regime did not ban congressional elections, but it did ban presidential elections from the mid-1960s to the mid-1980s. 44

53 authoritarian elites induce major media outlets to support the regime status quo and ignore or defame the opposition. In Taiwan during the early 1990s, the Kuomintang s (KMT) control of mass media prevented opposition victories in key elections. 156 Malaysia s United Malays National Organization (UMNO) has dominated national parliamentary elections since independence in 1957, due in part to the UMNO s ownership of major media corporations (Levitsky and Way 2010: 321). Under democracy, opposition parties do not face systematic media hostility. As a result, media-based parties tend to form in democratic contexts. Third, in some cases, authoritarian contexts give rise to high levels of polarization, conflict, and (less often) violence. Authoritarian regimes exclude and handicap political opponents and may employ violence against them. Political exclusion, hostility, and violence can galvanize the opposition and, especially when coupled with weak government performance, give rise to broad anti-regime movements. Since few regimes practice extreme, nationwide repression, authoritarian rule, at least in the long term, rarely prevents committed political opponents from organizing such movements. Thus, instead of diminishing prospects for successful opposition party-building, authoritarian contexts, by giving rise to mass mobilization, polarization, and conflict, often produce the higher causes and committed activists that are critical for party-building. When authoritarian regimes engage in violence, they may intensify selection pressures for activist commitment and generate opposition cohesion. Under democracy, the political opposition does not face systematic exclusion, defamation, coercion, or violence. Consequently, polarization, conflict, and violence do not often escalate to the levels associated with movements for democracy or struggles against authoritarian violence. 156 Media access...was badly skewed: All three television networks, most radio stations, and and Taiwan s two leading newspapers were in the hands of the state, the KMT, or allies (Levitksy and Way 2010: 315). 45

54 Low access to state resources Authoritarian rule Low access to mass media Organizational strength, activist commitment New party survival Polarization and conflict Figure 1.3: The role of authoritarianism In linking new party survival to conditions of adversity and (less strictly) authoritarianism, the book breaks with conventional wisdom on contemporary democratization, suggesting that the third wave has inhibited contemporary party-building rather than facilitating it. It should be emphasized, however, that authoritarian rule is not an independent variable. The section has simply identified a loose empirical association between a cluster of independent variables and the regime type, authoritarianism. After the formative phase This theory section has argued, among other things, that new parties are more likely to survive the formative years if they initially build strong organizations. One might expect such parties to become organizationally weaker over time, as leading party members oc- 46

55 cupy government positions, and the party gains traction in the broader electorate, rendering organization less electorally necessary. Due to three countervailing factors, however, most parties remain organizationally strong or increase their organizational strength after the formative phase. 157 First, even after a party gains broad traction in the electorate, organization-building continues to provide marginal electoral benefits that party elites may value, particularly if the elites in question prioritize lower-level elections. 158 Second, organizational maintenance and expansion become easier over time. Leaders and activists accumulate party-building experience, and electoral success strengthens party finances through increased state funding and outside contributions. Third, the initial period of organization-building may produce internal constituencies that benefit psychologically, socially, and materially from the party organization and demand its maintenance or expansion. 159 Brazil s PT, Mexico s PRD, Uruguay s FA, and El Salvador s FMLN each have expanded their organizations since the formative phase and continue to enjoy an organizational advantage in large regions of their countries. 160 Although organization contributes to party institutionalization, very few parties become stable, long-term contenders for national power due to organization alone. Most institutionalized parties combine strong organizations with strong brands. 161 As already emphasized, most parties, in order to develop a successful brand, must initially establish a distinct profile 157 In some cases, the primary purpose of the party organization may shift from mobilization to patronage. This shift has occurred, for example, within Argentina s PJ (Levitsky 2003). 158 See Van Dyck and Montero (n.d.); Van Dyck (n.d.). 159 Party leaders also might choose to continue investing because they have paid the sunk costs of infrastructural penetration. 160 On the PT, see Ribeiro (2010); Van Dyck (2014). 161 Neither a strong organization nor a strong label is necessary for party institutionalization. Some parties with weak organizations survive the formative phase and become institutionalized on the strength of their label alone (e.g., Chile s PPD), while some parties with large infrastructures survive in the long term without partisan supporters (e.g., Brazil s PMDB, a brandless patronage machine with a larger membership and more territorially extensive infrastructure than any other Brazilian party). 47

56 within their political system and maintain this profile over some period of time. Since it takes time to demonstrate consistency, party brands remain relatively weak during the formative years. Parties that survive the formative years, however, tend to consolidate brands, many of which last for decades. To the extent that a party possesses a strong brand, and hence a loyal partisan base in the broader electorate, elites and members will have an electoral incentive not to defect. Consequently, even though parties tend to lose their initial source/s of cohesion (e.g., violent struggle, a particular party leader) over time, the party brand typically becomes a key source of longer-term cohesion. With respect to durability, weak brands are the main factor that distinguish new parties from institutionalized parties and, other things equal, make the former more vulnerable to collapse. Case overviews, operationalization, and research design Chapters Two through Seven will present analytic narratives of two new left survivors, Brazil s Workers Party (PT) and Mexico s Party of the Democratic Revolution (PRD), and two cases of new left collapse, Argentina s Front for a Country in Solidarity (FREPASO) and Peru s United Left coalition (IU). This final section presents an overview of the arguments in Chapters Two through Seven and closes with a discussion of operationalization and research design. Case overviews Why did the PT and PRD survive, while FREPASO and the IU collapsed? The four cases subdivide into two potentially useful paired comparisons: the PT with the IU, and the PRD with FREPASO. 162 The PT and IU formed in 1980 as mass-based, left-wing, opposition 162 Abal Medina (1998) also notes the striking similarities between FREPASO and the PRD, detailed below. 48

57 forces bringing together a heterogeneous mix of radical and moderate left elements. Both were born into relatively weak party systems without a stable left party. Brazil and Peru had similar electoral systems, with presidentialism, full proportional representation, and (especially in Brazil s case) additional features considered inimical to the formation of stable parties. Both quickly established large activist bases and territorial infrastructures. The PT survived the formative years and today anchors Brazil s partisan left. The IU split apart and collapsed, defying scholarly predictions. To this day, Peru remains without a major left party. Mexico s PRD and Argentina s FREPASO also have parallel formations. Both originated as left-wing, anti-neoliberal, elite splinter groups from dominant, traditionally populist parties that had adopted market reforms and fiscal austerity: the PRI in Mexico and the Peronists (PJ) in Argentina. The schisms that gave rise to the parties both occurred around the time that the Soviet Union was collapsing: 1988 in Mexico, 1990 in Argentina. Both parties brought together a heterogeneous mix of radical and moderate left elements. Both were born into similar party systems, where two traditional parties had dominated electoral competition for many decades. 163 Both Mexico and Argentina had presidentialism with proportional representation, full (Argentina) or mixed-member (Mexico). During the 1990s, the PRD and FREPASO became the first new parties in decades to challenge the two-party establishments that had long prevailed in their countries. Both parties spent the bulk of their formative years with virtually no access to state resources. The PRD survived early development and, like Brazil s PT, today anchors Mexico s partisan left. FREPASO defied scholarly predictions by collapsing amid electoral crisis in the early 2000s. Argentina remains without a major left party to this day In Mexico, these two parties were the PRI and the PAN, the latter a conservative party that was fifty-years old. In Argentina, the two parties were the Peronist PJ and the Radical Civil Union (UCR), a centrist party more than a century old. 164 Leftist elements (e.g., kirchneristas), however, do periodically control the ideologically shifty PJ. 49

58 Figure 1.4: Electoral history in Chamber PT, PRD, IU, FREPASO The PT and PRD have each participated in eight congressional elections. The PT has met the ten-percent threshold in six consecutive elections, from 1990 to the present. The PRD has met the ten-percent threshold in seven consecutive elections, from 1994 to the present. In Peru, congressional elections are held only twice a decade. The IU, created just months after the 1980 congressional elections, only participated in two congressional elections, easily passing the ten-percent threshold in 1985, falling slightly below it in 1990, and breaking apart soon after. In Argentina, congressional elections are held every two years. FREPASO and its precursors (FREDEJUSO, FG) participated in five congressional elections, passing the ten-percent threshold in the middle three, suffering a crisis in the fifth election in 2001, and collapsing soon after. Paradoxically, the survival of the PT and PRD can be traced, in large measure, to early adversity. For both parties, adverse conditions during the formative period led to the creation 50

59 of strong party organizations composed of committed activists. Organizational strength and activist commitment enabled them to survive early electoral disappointments and crises. Specifically, both the PT and PRD spent the bulk of their first decades in the opposition, with few executive positions and hence meager access to state resources. In 1980s Brazil and 1990s Mexico, conservative opponents of the PT and PRD monopolized television and radio, leaving the PT and PRD with scarce access to media. Low access to state resources and media created an incentive for the leadership of both parties to invest in organization, and also selected for committed activists willing to build the PT and PRD from the ground up, without remuneration or prospects for a plum government position. The PT grew out of the unprecedented mass labor mobilizations of the late 1970s, the PRD out of the mass cardenista anti-fraud movement launched in response to the stolen 1988 presidential election. These formative struggles produced the higher causes that spurred base-level petistas and perredistas to collective action. Heterogeneous petistas united around the cause of empowering workers and the popular sectors in a historically elitist political system. Similarly heterogeneous perredistas united around the cause of democratizing Mexico and rolling back 1980s neoliberalism, both to be achieved by putting Cárdenas in the presidency. Both parties built their organizations on civil society structures: new unions, Catholic grassroots communities, and Marxist networks in the PT s case; the traditional Marxist left, the extraparliamentary social left, and defecting PRI structures in the PRD s case. Both parties, but especially the PRD, suffered early repression in select regions and localities, which hardened activist commitments. Organizational strength and activist commitment fortified both parties during early development. The PT and PRD suffered abysmal first elections (in 1982 and 1991, respectively) but rebounded due to the persistence of activist networks in territorial strongholds. The survival of the PT and PRD, however, cannot be traced exclusively to organizational 51

60 strength and activist commitment rooted in adversity. Given their origins in broad mass movements, both parties were left fronts with high levels of internal heterogeneity. Party leaders Lula da Silva (PT) and Cuauhtémoc Cárdenas (PRD) played decisive roles in preventing fatal schisms. Both leaders combined electoral coattails and unquestioned internal authority. Consequently, they furnished electoral incentives against defection and had the capacity to arbitrate and impose during critical moments of internal conflict. In this way, while the early PT and PRD were still attempting to build strong brands and institutionalize conflict resolution procedures, their leaders substituted for both. This chapter has argued that (1) a strong organization composed of committed activists and (2) conditions for avoiding schism (e.g., a strong party leader) are critical for new parties. In the broadest terms, FREPASO collapsed because it lacked the first, while the IU collapsed because it lacked the second. FREPASO s collapse, paradoxically, can be traced to its advantageous origins. Unlike the PT and PRD, FREPASO emerged and developed under democracy, in a relatively auspicious context. FREPASO spent the bulk of its first decade in the national opposition, with little access to state resources, but the party developed under conditions of press freedom, with regular and unfettered access to the country s major television, radio, and newspaper outlets. Although FREPASO did not emerge in a context of intense polarization or conflict, party elites did have ties to independent working-class and teachers unions that, like FREPASO, belonged to the political left and opposed the PJ. These unions might have provided the organization platform for a grassroots party-building effort, but FREPASO s leaders, the media-savvy Chacho Álvarez foremost among them, intentionally kept the unions at arm s length. Facing a relentless electoral calendar, the top FREPASO leadership calculated that an extensive partisan activist base would take time and labor to develop and, crucially, would 52

61 demand a radical, electorally suboptimal economic program. Álvarez and his closest associates preferred to target floating middle-class progressives with media appeals, and to tone down the party s previously strident economic rhetoric in favor of a clean government platform. For these reasons, the FREPASO cúpula completely eschewed organization-building. As FREPASO shot up in the polls, the majority of Argentina s left elites and groups including many who might have wished to build an Argentina version of Brazil s PT jumped on the FREPASO bandwagon. When FREPASO, as junior partner to the Radicals, won the vice presidency and a shared congressional majority in 1999, the party amounted to a media-based brand, property of a few dozen elites. Calamity struck in , and FREPASO was not organizationally equipped to survive. Amid a violent economic crisis and corruption scandal, Álvarez, in an effort to save the FREPASO brand, resigned the vice presidency in The gambit failed, and in the 2001 midterms, FREPASO lost sixty percent of its Chamber seats. The brand had stopped delivering, and FREPASO elites jumped ship. With no activist networks to fall back on, FREPASO collapsed in the months following the 2001 election. The IU, like the PT and PRD, was born in adversity and established a strong territorial organization with a large, committed activist base. After leading extended mass mobilizations against Peru s right-tacking Morales Bermúdez military government in the late 1970s, the country s left parties and associated organizations and movements united to form the IU in advance of the 1980 general election, which marked Peru s democratic transition. The IU remained in the national opposition throughout the 1980s, although coalition leader Alfonso Barrantes did hold the important Lima mayoralty from 1983 to Throughout the decade, the IU enjoyed little access to Peru s independent but concentrated and conservative television and newspaper establishments. Electorally, the IU depended on an extensive, primarily volunteer activist base, driven to halt the right s advances and return the left to 53

62 power. Although the IU inherited many of these activists from its constituent parties, a large number of independent IU supporters (the no partidarizados) also became active at the base level during the 1980s. The IU, in contrast to FREPASO, was organizationally equipped to endure. Despite this, the coalition suffered a fatal schism in This schism resulted from coalition leader Alfonso Barrantes s weak internal authority. IU leaders gave the top position to Barrantes, an independent ex-aprista, in part because he possessed electoral appeal, but also because he lacked ties to any of the coalition s founding parties and organizations, and hence did not threaten the internal balance of power between constituent parties. Although IU elites recognized, almost universally, that Barrantes stood the best chance of winning the Peruvian presidency, Barrantes, in contrast to the PT s Lula, the PRD s Cárdenas, did not possess unquestioned internal authority. He lacked ties to the IU s major elites and constituent parties and could not claim any special source of broad moral legitimacy within the coalition. Leaders of the IU s radical bloc (e.g., Javier Diez Canseco), who dominated the IU s internal organization, openly described Barrantes as a populist caudillo, not an authentic left leader. In the deft, unprompted formulation of one moderate coalition member, Barrantes was accepted as a candidate but disputed (discutido) as a leader. 165 From the IU s inception, Barrantes lacked internal decision-making authority in critical zones of uncertainty. In contrast to the PRD s Cárdenas, who unilaterally vetoed early calls for a unanimity requirement in the PRD s National Executive Committee, and who imposed himself as presidential candidate in 1994 and 2000 despite significant internal competition, Barrantes tried and failed (1) to veto the unanimity requirement in the IU s National Executive Committee early in the 1980s, (2) to tame IU radicals on the internally polarizing issues of the late 1980s (e.g., the IU s stance on the Shining Path and Alan García), and (3) 165 Author s interview with Panfichi. 54

63 to impose himself as the IU s presidential candidate in In the lead-up to the 1990 presidential election, IU radicals refused to grant Barrantes the presidential nomination without a primary, and Barrantes defected to run independently. With the previously united left split down the middle, both Barrantes and the IU performed abysmally in the 1990 election, and the IU dissolved shortly thereafter. The divergent origins and outcomes of the PT, the PRD, FREPASO, and the IU fit a broader trend within the new left. By and large, Latin America s surviving new left parties and coalitions developed in adversity. Specifically, most new left survivors spent key formative years in opposition to authoritarian rule, either as new opposition parties (e.g., PT, PRD) or insurgent precursors engaged in civil war (FMLN, FSLN). The PT spent its first five years (1980-5) under military-authoritarian rule. The PRD spent its first eleven years under PRI hegemony ( ). As guerrilla armies, the FSLN and FMLN spent at least a decade under authoritarian rule. The out-of-sample left survivor, Uruguay s FA (est. 1971), spent eleven of its first thirteen years under military rule ( ). In contrast, every major new left failure was born under democracy. Table 1.2: Authoritarian origins and new left survival Survival Collapse Authoritarianism FSLN, FMLN, PT, PPD, PRD Democracy MAS IU, M19, PEN, FREPASO, MAS, FDNG 55

64 Earlier, the chapter raised the question: Why, given the various obstacles facing the new left in Latin America, did new left parties fare well relative to their right-wing counterparts? This book points to a possible answer. New left parties fared relatively well after the onset of the third wave because, more often than new right parties, they emerged from adversity. Operationalization In preview of the analytic narratives presented in the empirical chapters, this section provides scores for each of the book s four main cases on the theory s independent and intermediate variables. As detailed earlier, the determinants of organizational strength and activist commitment include access to state resources, access to mass media, access to mobilizing structures, and origins in polarization and conflict. For these four independent variables, the parties receive a score based on the following questions, respectively: Did the party win a major executive position within five years? Were mass media independent and free of systematic conservative bias during the party s first five years? Did party elites have ties to large mobilizing structures during the first five years? Did the party emerge from intense societal polarization and conflict? The questions have a relatively short, five-year time frame because office-seekers incentives and capabilities in the initial few years almost invariably determine whether leaders will invest in, and succeed in building, a solid, durable organization. Winning a major mayoralty or governorship during this time does not provide a level of state access sufficient to substitute for a national party-building effort (although it may be useful in other ways). Parties that wish to become national electoral forces, as all major new left parties did, need electoral capacity across the country, not just in a few large cities or a few states. Only the national government apparatus can substitute for a national party. Thus, parties that won the presidency within five years receive a full score for state access, while those that won a major subnational position receive a limited score, and the rest receive a none score. 56

65 Along similar lines, for the mobilizing structures variable, parties receive a full score if, during the first five years, party elites had ties to preexisting organizations and movements with national mass memberships (1 member per 1000 citizens), as well as infrastructure in one fifth of the country s municipalities and, where applicable, in half of its states or provinces. If party elites had ties to mobilizing structures without national reach, but with a major presence in one or a few big cities, or in most of the municipalities of one or two states, they receive a limited score. If a party s elites did not have ties to mobilizing structures with national or subnational reach, they receive a none score. New left parties that initially developed in a context of press independence did not necessarily enjoy a level of media access sufficient for building a national electoral constituency. In countries where mass media were independent but concentrated in the hands of conservative forces, new left parties could not make effective national appeals. Consequently, cases only receive a full score on media access if, during the first five years, mass media were independent and free of systematic conservative bias. If mass media were independent but systematically conservative, new left parties receive a limited score, and if mass media were not independent, parties receive a none score. Finally, for the origins in polarization and conflict variable, the book distinguishes between violent conflict (e.g., civil war) and nonviolent conflict (e.g., peaceful, anti-regime mobilizations). Although both types of conflict generate large-scale activist causes, violent conflict produces additional effects on party-building, namely selection pressures and intragroup dynamics that help prevent fragmentation and schism. If a new party does not emerge from either type of struggle, it receives a no conflict score. This section also scores the cases on a composite of the intermediate variables, organizational strength and activist commitment. A party receives a yes score for this composite variable if, in the first five years, two conditions held: (1) the ratio of party members to population 57

66 reached 1 per 1000, and the party established formal branches in twenty percent of the nation s municipalities and, where applicable, half of the states (organizational strength); (2) the party depended primarily on unpaid volunteers for grassroots party work (activist commitment). Table 1.3: Organization and activist commitment Party PT (Brazil) PRD (Mexico) IU (Peru) FREPASO (Argentina) State access Media access Mobilizing structures Polarization/ conflict Organization with committed activists None None Full Nonviolent Yes conflict None None Full Nonviolent Yes conflict Limited Limited Full Nonviolent Yes conflict None Full Full No conflict No Table 1.4: Survival and collapse Party PT (Brazil) PRD (Mexico) IU (Peru) FREPASO (Argentina) Organization with Leader with coattails committed activists and internal authority Outcome Yes Yes Survival Yes Yes Survival Yes No Collapse by schism No Yes Electoral collapse 58

67 Finally, the section scores each of the four cases on a composite of the party leader s coattails and internal authority. Party leaders receive a yes score if, during the first five years, two conditions held: (1) most major party elites believed that the leader, compared to all other elites, had the best chance of victory in a top executive election (coattails); (2) the leader held a unique position at the hub of the party s network of relationships or possessed a special source of credibility, respect, or moral legitimacy within the party (internal authority). Research design Party-building remains an undertheorized area of research despite the high volume of existing literature on parties. Thus, the book s empirical research design is mainly geared toward theory generation. Chapters Two through Eight nest four in-depth case studies in a broader, medium-n comparison. The case studies do not test the book s theory. Rather, they function as plausibility probes (Eckstein 1975), evidencing the theory s initial plausibility by illustrating its key causal mechanisms at work in a few select cases. In-depth, small-n research enables the tracing of micro-level causal mechanisms and thus facilitates the generation of robust theories (Hall 2007). The conclusion, Chapter Eight, provides evidence of the theory s generalizability, supplementing the four case studies with an analysis, based on desk research, of the remaining new left left survivors and failures and several out-of-sample cases (e.g., Uruguay s FA, Venezuela s LCR). The book s case universe is restricted to major new left parties/coalitions on methodological grounds. In order to understand the conditions for successful party-building, one should, to the extent possible, compare cases of survival and collapse that differ as little as possible on analytically relevant dimensions, observed and unobserved. Latin American countries are characterized by broad structural and institutional similarities, which include religion (Catholicism), socioeconomic measures (middle-income levels, high inequality), and electoral rules (presidentialism, PR). The region s major new left parties and coalitions bear additional 59

68 similarities. As detailed earlier, these parties arose in the same time period ( ) and had to adapt to the collapse of communism, the neoliberal turn, the decline of unions, and the expansion of the atomized informal working class. Despite these and other similarities, Latin America s new left parties vary on the outcome of interest, survival. Thus, they may be treated as most similar cases. Chapters Two through Seven draw on data from months of interviews and archival research in São Paulo, Campinas, and Brasília, Brazil; Mexico City, Mexico; Lima, Peru; and Buenos Aires, Argentina. I conducted eighty-five interviews with party elites, party activists, party observers, and country-based scholars and examined over 1000 documents from official party archives, newspaper archives, and interviewees personal archives. The empirical chapters draw on numerous additional sources, including dozens of published interviews with party members, 166 retrospective first-hand testimonies, 167 and detailed scholarly analyses conducted by party members. 168 To score the cases on independent and intervening variables, the chapters draw heavily on expert country analyses, especially first-hand accounts and analyses, as well as interviews (especially for the leadership variables). The most useful sources included Meneguello (1989), Keck (1992), Ribeiro (2010), Secco (2011), and Rodríguez (n.d.) for the PT; Bruhn (1998), Borjas (2003), Martínez (2005), and Rodríguez (n.d.) for the PRD; Novaro and Palermo (1998) and Abal Medina (2009) for FREPASO; and Cameron (1994), Roberts (1998), Herrera (2002), and Adrianzén (ed.) (2011) for the IU. While first-hand accounts/analyses and secondary materials furnished most of the evidence necessary for scoring variables, interviews and archives provided most of the evidence neces- 166 E.g., de Moraes and Fortes, eds. (2008); González et al, eds. (2010). 167 E.g., Herrera (2002). 168 Borjas (2003); Novaro and Palermo (1998); Secco (2011). 60

69 sary for demonstrating causal mechanisms, or the relationships between variables. Interviews present methodological challenges, as ideas and events that may not have seemed important to participating actors in the moment (e.g., a party s reliance on mass media or quick access to state resources) might come to seem important in retrospect. For this reason, it is crucial, in interviews, to pose general, open-ended questions and listen for unprompted statements, in order to avoid implanting ideas in the interviewee s mind. 169 Yet the validity of interview evidence also depends on repetition across a diverse range of interviews. Insofar as multiple interviewees with different beliefs and loyalties (e.g., radicals v. moderates in a particular party) independently confirm that a particular meeting occurred, or that a particular line of thought prevailed among the party elite or activist base, one can be confident that the interviewees have not confabulated or provided idiosyncratic, unrepresentative interpretations of events. 170 Still, no interviewing technique can wholly circumvent the problem of hindsight. Thus, archives are an invaluable source of field evidence. Unlike interviews, archives reveal what participants and observers thought and expressed in real time, before they knew how events would ultimately unfold. 171 The PT chapters draw extensively on materials from the Perseu 169 As an illustration, in Mexico, PRI operatives murdered hundreds of PRD activists during the late 1980s and early to mid-1990s, and scholars such as Greene (2007) and Rodríguez (n.d.) have posited, with evidence, that these repressive conditions selected for ideological activists and elites. Yet no existing studies explore the collective psychological effects of PRI violence on local networks of perredistas. When asked to describe the PRD s reaction to the electoral crises of 1991 and 1994, one party activist from San Luis Potosí, without any previous mention of PRI violence, stated, we were at war. This provided a small bit of evidence that, at the base level, PRI hostility and violence sharpened PRI/PRD boundaries and hardened perredistas collective identity. 170 If, for example, only a lone, radical interviewee had highlighted the toxic relationship between Alfonso Barrantes and IU radicals, Chapter Seven could not, with justification, have characterized Barrantes crossfactional ties as weak. Yet since several interviewees offered the same assessment, and most of them generally supported Barrantes, the characterization has a relatively solid basis. 171 Official PT archives from the 1980s, for example, provide in-the-moment evidence that low access to media contributed to organization-building, as party organizers, in their campaign manuals and pamphlets, exhorted activists to donate time and resources given the PT s resource disadvantage and the opposition of the conservative media establishment (Chapter Two). 61

70 Abramo Foundation s Sérgio Buarque de Holanda Center: Documentation and Political Memory, and from the Edgar Leuenroth Archive at the University of Campinas. The PRD chapters rely heavily on articles published in the Mexican weekly magazine, Proceso. All Proceso excerpts were originally cited in Borjas (2003) unless otherwise noted. The FREPASO chapter makes extensive use of Marcos Novaro s valuable personal archive. Finally, the Peru chapter draws heavily on articles published in the Peruvian monthly magazine, Quehacer. * * * The rest of the book is organized as follows. Chapters Two and Three analyze the survival of Brazil s PT, while Chapters Four and Five analyze the survival of Mexico s PRD. The two pairs of chapters have the same analytical structure. Chapters Two and Four show that for the PT (Chapter Two) and PRD (Chapter Four), early adversity led to the creation of a strong party organization composed of committed activists, and that organization and activist commitment enabled the survival of early electoral crisis. Chapters Three and Five argue that by combining coattails and unquestioned internal authority, Lula da Silva (Chapter Three) and Cuauhtémoc Cárdenas (Chapter Five) substituted for strong brands and institutionalized conflict resolution procedures and thus prevented fragmentation within the early PT and PRD. Chapters Six and Seven analyze the collapses of Argentina s FREPASO and Peru s IU, respectively. Chapter Six shows that FREPASO s elite founders rose to national prominence through mass media appeals, did not have an incentive to invest in grassroots organization, and were not organizationally equipped to survive the electoral crisis of the early 2000s. Chapter Seven argues that within the IU, Alfonso Barrantes furnished coattails but did not possess unquestioned authority and thus could not veto constraining decision rules, arbitrate internal conflict, or ultimately even impose himself as candidate. These facts, the chapter shows, prompted his exit and the IU s fatal schism. 62

71 The conclusion, Chapter Eight, summarizes the book s argument, explicitly compares the PT, PRD, FREPASO, and IU to each other, applies the book s argument to the broader set of new left cases and a few shadow cases, and discusses research implications. 63

72 Chapters 2-3 The Survival of Brazil s Workers Party (PT) Chapters Two and Three analyze the survival of Brazil s Workers Party (PT). Founded in 1980, the PT has contended for national power in Brazil since the late 1980s. Petistas have held ten governorships and nearly two-thousand mayoralties, including São Paulo s. The party has elected at least ten percent of Brazil s federal deputies in every congressional election since In recent years, the PT has consolidated its position as Brazil s leading party, retaining the presidency in 2006 and 2010 after party leader Lula da Silva s first successful presidential bid in Given Brazil s size, economic strength, and growing presence on the global stage, the PT is not just the most important new party in Latin America, but among the most important in the world. Figure 2.1: The PT in the Chamber of Deputies (1982-present) Dozens of scholars have studied the PT in depth. The bulk of existing literature either examines why the PT successfully adapted, moderated, and reached the presidency during 64

73 the 1990s and 2000s, or investigates the causes and effects of well-known PT institutions, governing practices, and policies such as early internal democracy, participatory budgeting in municipal government, and the Bolsa Família conditional cash transfer program. Chapters Two and Three do not directly concern any of these issues. Instead of asking why the PT adapted and reached the presidency, or why the PT has adopted certain distinctive practices and policies, the chapters ask a more basic question: Why did the PT survive and become institutionalized as a major national party in Brazil? 65

74 Chapter 2 Building the PT: adversity and survival The weapons of power are armies, money, and information. Slogan on PT flier, undated 1 The PT survived because of early adversity, not in spite of it. The PT originated in mass struggle against Brazil s military regime ( ). The PT s founders belonged to and drew from three social forces that played central, collaborative roles in this struggle: the organized Marxist left, the grassroots Catholic left, and above all the autonomous new union movement. The new unions and their civil society allies on the Catholic and Marxist left conceived the PT as a partisan vehicle for democratic and socialist transformation. In order to effect change in the institutional sphere, the PT, at a minimum, would have to achieve electoral success at the national level. Yet the early PT lacked access to finance and patronage. With no ties to the economic elite, and with federal laws banning union donations, the PT, from inception, would fundamentally depend on public resources. Public resources, however, proved scarce. Not until 1989, when Luiza Erundina assumed the São Paulo mayoralty, did the PT occupy its first major executive position. The early PT also lacked access to media. Under military rule and the early Nova República, the PT confronted a powerful mass media establishment dominated by the Globo network and both closed and opposed to the left opposition. Major news outlets denied the PT access, circulated open propaganda, and systematically omitted and edited news coverage to the PT s detriment. 1 From Edgar Leuenroth archive. 66

75 Low access to state resources and media created an incentive for PT elites to invest in organization. Unable to reach voters through media or attract activists and supporters with salaries, petty clientelism, and promises of patronage, the PT could only contend for national power if it developed a large, volunteer organization and, through it, distributed collective incentives to the broader population. PT leaders understood this. In campaign materials and internal strategic discussions, party leaders repeatedly emphasized, to each other and to the rank-and-file, that they depended on activists to deliver the unfiltered party message to would-be supporters. Elite incentives were insufficient for the creation of a strong PT organization. Party founders also had to be capable of large-scale organization-building. The PT s capacity for organizationbuilding came from founders origins, linkages, and support in civil and political society, primarily the new unionism, Catholic left, and Marxist left. Lacking the resources and institutional capacity to finance or administer large-scale territorial implantation from the top down, the national PT organization depended on the bottom-up initiative and dispersed human, material, and infrastructural resources of these civil society feeder organizations. Thus, low access to material resources and media made organization-building electorally imperative, while the PT leadership s origins and linkages in civil society made organizationbuilding possible. Yet the PT depended on one more vital ingredient: activist commitment based on belief in a higher cause. Individuals do not engage in time-consuming, labor-intensive, unremunerated, and sometimes risky partisan activism unless driven by a higher cause. The PT s higher cause went beyond opposition to military rule. Leaders branded the PT, and the rank-and-file viewed the PT, as a novelty in Brazilian history: the first authentically popular expression in a society perpetually dominated by elites. It was this narrative of popular struggle, which Lula and other new unionist party leaders embodied, that inspired thousands of local civil 67

76 society leaders and activists to build the PT organization from the ground up. During its formative period, the PT far surpassed the organizational requirements imposed by law. Petistas built the most internally vibrant of Brazil s major parties, with the largest volunteer activist base and one of the largest territorial infrastructures. Despite slow early growth and major initial setbacks (e.g., the 1982 elections), local activist networks in the PT s core bastions persisted, rebounded, and, together with party leaders, continued to invest in PT-building, new union-building, and the broadening of PT/civil society linkages. Organizational strength and activist commitment thus fortified the PT, enabling the party to survive a slow, often disappointing formative phase. By the late 1980s, the PT had consolidated Brazil s dominant partisan brand on the left and become institutionalized as a major national party. Setting the stage: military rule and the genesis of the Workers Party The Brazilian military seized power in 1964 with economic growth slowing, inflation and debt rising, worker and peasant mobilization intensifying, and within the military fear of Marxist insurgency mounting due to the radicalization of organized labor, the formation of urban and rural guerrilla groups within Brazil, and the recent consolidation of Marxist rule in Cuba. 2 National security hard-liners and economic statists controlled Brazil s military regime for most of the first decade of military rule ( ). Early regime technocrats viewed economic progress and social order as mutually reinforcing (Stepan 1971, 1973 (ed.); Moreira Alves 2 In their forcible takeover, Brazil s military golpistas enjoyed support from key societal and political actors/blocs, including most of the business class and rural landowning elite, numerous governors, much of the national media establishment, and segments of the middle class that prioritized the restoration of social order and resolution of Brazil s fiscal and economic troubles over democratic stability. 68

77 1988; Skidmore 1988). While social order would provide a secure platform for economic growth, broad-based material progress would mitigate and preempt social unrest, violent opposition, and other forms of instability. In order to restore and sustain social order, early military governments adopted a multipronged approach, restricting political contestation and protest, repressing left subversives, and actively attempting to unify the general public around regime ideology. The first military government of Castelo Branco (1964-7), by executive decree, imposed the basic authoritarian system that would last until the early 1980s. Shortly after inauguration, the government expelled influential opposition figures from political life 3 and canceled the 1965 presidential elections. 4 Following a strong opposition showing in the direct gubernatorial elections of 1965, 5 the Castelo Branco government banned all existing parties and created an official two-party system composed of the situacionista (ruling) ARENA and oposicionista Brazilian Democracy Movement (MDB). 6 ARENA and the MDB would remain Brazil s only legal parties until the liberalization of the party system in The country s two most hard-line military governments under Presidents Costa e Silva (1967-9) and Médici ( ) employed heavy repression in order to defeat subversives. From the late-1968 passage of the Fifth Institutional Act 7 until the mid-1970s a period 3 These included former president Juscelino Kubitschek. 4 The Castelo Branco government also intervened in hundreds of unions, banning strikes. 5 Opposition parties won in Minas Gerais and Guanabara, a former state encompassing modern-day Rio de Janeiro. 6 The Castelo Branco government also circumscribed the functions of the national legislature, expanded the decree-making powers of the presidency, and made the 1970 gubernatorial elections indirect. (Because the constitution stipulated direct gubernatorial elections, military governments could only impose indirect elections via executive decree.) The Castelo Branco government also outlawed direct elections of mayors, authorizing governors to appoint mayors instead. Government lawyers would incorporate many of these reforms into the military s 1967 Constitution. 7 Costa e Silva s term (1967-9) gave rise to the first serious outbreak of hostilities between the military regime and its opponents, but national security took center stage during the Médici years ( ). During Costa e Silva s term, the Revolutionary Popular Vanguard (VPR) perpetrated a fatal, military-directed 69

78 known as the Anos de Chumbo, or Lead Years Brazilian military and police murdered or disappeared between 100 and 500 individuals and committed other forms of repression on a much larger scale, arresting approximately 50,000 individuals, torturing 20,000, exiling 10,000, and suspending the political rights of (cassando) 5000 (Chiavenato 1997: 131; Keck 1992: 52). 8 Both governments primarily targeted the Marxist left, 9 which encompassed urban guerilla cells (responsible for dozens of bank robberies and kidnappings), 10 rural guerrilla forces, 11 clandestine parties and political organizations, and radical intellectuals and opinion-makers. From the hard-line military perspective, the Marxist left posed a range of threats to national security, from terrorism and armed revolution to the radicalization, infiltration, and mobilization of civil society, especially organized labor. In order to win the battle of ideas, Brazil s military rulers attempted, from the beginning, to unify the general public around regime ideals and goals in Amaral and Guimarães (1994) words, to guarantee national unity and the massification of the ideology of national security car bomb attack, and the police killed a student protester in Rio de Janeiro, triggering a mass protest in Rio, the (Passeata Cem Mil). Amid growing opposition to state repression among MDB legislators (e.g., Márcio Moreira Alves), Costa e Silva s National Security Council penned the Fifth Institutional Act (AI-5), empowering the federal executive to close the national legislature, remove elected officials, and use repression to limit political forms of expression and association. Following AI-5 s passage in December of 1968, the Costa e Silva government tightened authoritarian controls and began to create a culture of fear through coercion and violence. The administration closed Brazil s congress for nearly a year, passed and enforced strict censorship laws, and began to target subversives with a range of violent and repressive measures, from exile and suspensions of political rights (cassações) to detentions, torture, and disappearances (Moreira Alves 1988). 8 The Brazilian Human Rights Commission and Amnesty Commission have estimated 333 and 457, respectively. Other sources estimate between 100 and 200 (CITATION NEEDED). 9 In their use of cassações and forced exile, hard-line governments also targeted less radical individuals considered dangerous due to their social or political influence. These included political and cultural elites such as Juscelino Kubitschek, Gilberto Gil, Caetano Veloso, and many more. 10 Several of Brazil s most well-known urban guerrillas, including Carlos Marighella and Carlos Lamarca, died in armed exchanges with police in the late 1960s and early 1970s. 11 In , the Brazilian military engaged and defeated the rural guerrilla force, the Guerrilha do Araguaia (Pará), resulting in scores of casualties (CITATION). The Guerrilha s leaders modeled the force on the Chinese Revolution and had close ties to the PCdoB (Rodríguez n.d.: 142). With approximately eighty members, the Guerrilha was the strongest domestic insurgency in Brazil during the Cold War period. 70

79 (32). In pursuit of national integration and cultural homogenization, early military governments created Brazil s first public telecommunications companies, Embratel (est. 1965) and Telebrás (est. 1972), 12 and actively promoted the monopolization of Brazil s private media market, singling out the pro-regime TV Globo for special favor in the all-important, fast-growing television sector (Amaral and Guimarães 1994: 32). In 1970, the Médici government canceled the broadcasting license of the Excelsior television network, which had opposed the 1964 coup. The most overt attempts at top-down ideological indoctrination came in the early 1970s, when the Médici government sought to popularize nationalist, anti-opposition slogans such as Brazil: Love it or Leave it and If you don t live to serve Brazil, you don t serve to live in Brazil. 13 The federal government produced and circulated hundreds of thousands of bumper stickers with such slogans, while business allies placed the slogans on their products, and governista media networks (most importantly TV Globo) broadcast them. 14 In the economic domain, Brazil s hard-line military governments, in contrast to their Chilean and Argentine counterparts, systematically deepened the state s role in private and public industrial production (Skidmore 1988) and made massive, long-term investments in infrastructure, higher education, and the public safety net (Weyland 1994). After the Castelo Branco government (1964-7) stabilized the Brazilian economy, improved the fiscal outlook, and neutralized organized labor, 15 the successor governments of Costa e Silva and Médici, 12 The Cardoso administration would partially privatize Embratel and Telebràs in Brasil: Ame-o ou Deixe-o and Quem não vive para servir o Brasil, não serve para viver no Brasil. 14 CITATION NEEDED. 15 In order to curb inflation and ballooning public debt, which had partially motivated the 1964 military coup, the Castelo Branco government passed a set of orthodox stabilization measures, issuing a new currency (the cruzeiro novo), imposing strict monetary controls, implementing tax reforms designed to increase revenue, selectively cutting the budget and downsizing the public sector, and placing a set of constraints on organized labor intended to increase the efficiency of public and private corporations (e.g., bureaucratic reorganization, wage compressions, removal of employment guarantees). With respect to 71

80 under the stewardship of powerful finance minister Delfim Netto ( ), aggressively promoted infrastructural and human capital development and doubled down on public and private domestic production in sectors ranging from natural resource extraction to complex, high-end manufacturing. These governments ramped up funding for university research, provided critical financial support to fledgling private industries, poured resources into core primary-sector public companies (Petrobrás, Vale do Rio Doce), and seeking to move up the value chain founded ambitious new public companies such as Embraer (aerospace/defense, est. 1969). 16 In the early 1970s, the Médici government undertook a series of pharaonic infrastructural projects, including the construction of the Itaipú hydroelectric dam, Transamazônica highway, and the country s first nuclear plant in Angra dos Reis, Rio de Janeiro. From the late 1960s to the mid-1970s, the Brazilian economy grew at a dizzying pace. The Costa e Silva years (1967-9) saw a near two-fold increase in the rate of economic growth. The Médici years ( ) brought an economic miracle in which national income grew at an average of roughly ten percent, and tens of millions of Brazilians entered the middle class. Aided by increased tax revenues, the combined product of robust growth and Castelo Branco-era tax reforms, the Médici government implemented the most significant social policy expansions since the Vargas era, investing heavily in Brazil s higher education system and, in 1971, extending pension coverage to rural workers and domestic servants, two important and previously excluded occupational categories (Weyland 1994). In , however, Brazil s economy slowed and ground to a halt. In 1972, the Médici bureaucratic reorganization, in 1966 the Castelo Branco government, as part of an effort to neutralize organized labor, replaced the previous patchwork of sector-specific administrative agencies (IAPs) with a single administrative body, the INPS. The government also implemented modest industrial reforms, increasing support for public and private domestic industries and providing financial incentives for foreign industries to invest in Brazil. 16 Embraer was privatized in

81 government postponed direct gubernatorial elections until 1978, [foreseeing] probable defeat in important states in the 1974 election (Skidmore 1988: 150). No longer bolstered by a boom, the hard-line military elite began to lose support, especially among the middle sectors opposed to the repression, censorship, and electoral exclusion of the Lead Years. Consequently, the military s internal balance of power shifted toward soft-liners. The March 1974 inauguration of Brazil s fourth military government, headed by Petrobrás president General Ernesto Geisel, ushered in a less repressive, authoritarian phase of military rule. The Geisel government initiated a slow, gradual, and safe political opening (abertura), defined by an easing of repression and electoral exclusion (distensão). The Geisel government allowed free electoral advertisements on television and radio and began an attempt to co-opt civil society opponents into the institutional sphere (von Mettenheim 1995: 100). In the November 1974 congressional election, however, the MDB surged and nearly defeated the ruling ARENA, provoking a short-term hard-line retrenchment. 17 In , the Geisel government escalated its use of repression. Detentions spiked, new disappearances came to light, and police committed two notorious, high-profile murders in São Paulo, killing opposition journalist Vladimir Herzog (October 1975) and autonomous metalworker Manoel Fiel Filho (January 1976). 18 The Geisel government also formally restricted political parties media access. The Lei Falcão, passed in July 1976, prohibited political debates/advertisements on television/radio and only allowed parties, during official campaign season, to display photos of their candidates (Secco 2011: 81). Authoritarian retrenchment did not serve the military s goals. Political and societal opposition continued to intensify, such that by early 1977, the Geisel government foresaw major potential losses in the direct gubernatorial elections of While dialing down its use 17 The MDB s seat share increased from eighteen percent to thirty percent in the Senate, and from twenty-two to forty-four percent in the Chamber (compared to 1970). 18 von Mettenheim 1995:

82 of hard repression, the Geisel government closed the national congress in early 1977 and passed the April Package (Pacote de Abril), a set of executive decrees making the 1978 gubernatorial elections indirect, strengthening ARENA s presence in the national congress, and extending presidential terms to six years. Due to a combination of long- and short-term factors, opposition to military rule, by the late 1970s, had broadened and coalesced into a national movement (Moreira Alves 1988). Despite renewed economic growth and social policy expansion under Geisel, 19 tens of millions of Brazilians had not participated meaningfully in, or benefited significantly from, the country s material advances under military rule. In fact, poverty and inequality had deepened. Lowerclass wages had not kept apace with inflation, leading to reductions in the real income levels of the urban poor. Vertiginous economic growth at the top had widened the notorious chasm between Brazil s upper and lower classes. Rapid urbanization and the deepening of poverty had exacerbated problems of unemployment, crime, malnutrition, overcrowding, and inadequate infrastructure in the favelas of major cities (e.g., São Paulo, Rio de Janeiro). The problems of poverty, inequality, and urban dysfunction, coupled with the memory of the Médici years, recent galvanizing incidents such as the Herzog and Fiel Filho murders, and the continued denial of basic democratic freedoms, provided ideological fuel for the regime opposition as the second half of the 1970s progressed. Equally, the diminution and eventual elimination of hard repression under Geisel lowered the cost of open dissent, causing many exiles to return to Brazil and leading relatively moderate or risk-averse individuals to participate in the growing opposition movement. The balance of power between regime and opposition only shifted decisively, however, with 19 Robust economic growth persisted under Geisel, such that by 1980, the Brazilian economy had more than tripled in size since 1967 (in inflation-adjusted 1980 reais ( The Geisel government also implemented a new set of broad-based social policies, including progressive tax reform, the universal extension of emergency health care, and the extension of basic health care to the rural poor. 74

83 the rise of Brazil s autonomous union movement. Industrial and economic development during the 1960s and 1970s brought a major increase in the size of Brazil s public and private workforces, across the primary, industrial, and service sectors. From 1960 to 1978, the number of unionized workers shot from 1.5 million to over 10 million (Keck 1992: 13). Because the late Médici and Geisel governments rolled back some of the most draconian wage and employment tenure policies of the hard-line era, Brazilian organized labor, even the more autonomous sectors, initially remained on the sidelines of the opposition movement. Yet in late 1977, a World Bank study revealed that the Médici government had doctored inflation figures in 1973, dramatically underreporting the rising cost of food, housing, and other core goods and services in order to keep workers indexed wages low in real terms. 20 The revelation, publicly confirmed by the Geisel government, led to mass labor mobilization among a set of autonomous unions (os autênticos) in the industrial belt of Greater São Paulo, known as the ABC (or ABCD) paulista. 21 The São Bernardo Metalworkers Union, headed by Lula da Silva (union secretary since 1972), led the nascent movement (Sluyter-Beltrão 2010). Since the early 1970s, os autênticos had explicitly rejected Brazil s longstanding corporatist system of state/labor relations. Under this system, the Brazilian political elite, since the Estado Novo period ( ), had prohibited cross-sectoral unionization, instead engaging unions vertically by individual sector and, according to autonomous unionists, cowing and coopting union leaders (derisively called pêlegos) at the expense of rank-and-file workers. This pattern of relations had persisted, with minor variations, across authoritarian and democratic periods. 20 In 1973, due to the global oil supply shock and resulting spike in inflation, the Médici government repositioned worker salaries by indexing wages to inflation, across the board. 21 The ABC region comprises the outlying cities of Santo Amaro, São Bernardo do Campo, São Caetano, and (later) Diadêma and housed the core and supporting industries associated with Brazil s booming automobile sector. With the addition of Diadêma (formerly part of São Bernardo do Campo, some began to call the ABC paulista the ABCD paulista. 75

84 The wage recovery campaign thrust the autonomous union movement into the national spotlight. The new unionism (novo sindicalismo), as it came to be called, rapidly spread to urban centers and rural areas across regions and occupational sectors. By 1978, the movement counted nearly three million workers, roughly a quarter of Brazil s formal-sector working class (Rodríguez n.d.: 145). 22 In 1978, the metalworkers of the ABC paulista spearheaded the first in a wave of general strikes, pressing for higher real wages as well as more favorable tenure policies, increased shop-floor representation, and relaxed criteria for union formation and expansion. As the new unionism s star rose in the late 1970s, the Catholic left became heavily involved in the autonomous labor struggle. The progressive wing of the Catholic Church had emerged as a key opposition force during the hard-line era of military rule. 23 Beginning in the early 1960s, Brazil had witnessed a proliferation of Catholic grassroots communities (CEBs): small groups of lay Catholics, usually between ten and seventy individuals, who organized on a regular basis, in or outside of church, to analyze the Bible, examine the relationship between biblical teachings and everyday experience, and discuss political and social issues generally (Cavendish 1994). The CEBs emerged in a broader context of Church liberalization and reorientation toward issues of poverty and social justice, reflected globally in the Second Vatican Council ( ) and regionally in the rise of Liberation Theology (Gutiérrez 1971) and the Catholic Conferences of Medellín (1968) and Puebla (1979). From the early 1960s to mid-1970s, tens of thousands of CEBs sprouted up in rural areas, middle-class urban neighborhoods, and slums (e.g., the favelas encircling metropolitan São Paulo). By the end of the Anos de Chumbo, Brazil counted approximately 40,000 CEBs (Skidmore 1988). As authoritarian constraints tightened and repression increased under 22 This made the New Unionism the largest independent labor movement in Brazilian history. 23 Levine 1986; Mainwaring 1986; Skidmore 1988; Cavendish

85 Costa e Silva and Médici, the Catholic Church provided an umbrella of safety for regime opponents (Cavendish 1994: 187). CEBs became one of a few secure fora for political dissent in the country. 24 During the harshest years of repression, CEBs and interlinked neighborhood associations became hubs of opposition activity, giving refuge to potential targets of repression and offering resources and safe meeting places to dissident groups of various stripes (Cavendish 1994). CEB members affinity with the new unionism stemmed from a range of factors, including common class backgrounds (CEB members belonged overwhelmingly to the popular classes ), similar organizational cultures (e.g., grassroots participation), and a range of shared values (e.g., equality, anticentralism, bottom-up change). CEBs supported and encouraged the formation of rural and industrial new unions through the Pastorais da Terra and Pastorais Operárias. CEBs also built bridges between the new unions and other social movements, particularly poor neighborhood associations (Keck 1992: 79). Minority sectors of Brazil s Marxist left also became involved in the autonomous labor struggle. Despite competition and conflict between the new unionism and the communist party establishment (i.e., PCB, PCdoB, MR-8), 25 numerous smaller Marxist networks and organizations, drawn to the new unionism s militant class consciousness, created new union cells to foment and participate in the autonomous mobilization of organized labor. As the labor movement s external ties proliferated, new unionist demands and goals broadened from labor reform to democratization and the reduction of poverty and inequality through socialist policies. [T]he new unionism soon established itself at the forefront of the democratic opposition forces (Sluyter-Beltrão 2010: 4). 24 Levine 1986; Mainwaring 1986; Skidmore 1988; Cavendish About which more below. 77

86 In response to the general strike of 1978, the Geisel government conceded real wage increases but did not give ground in other areas. Consequently, a view took hold within the new unions that autonomous workers needed political representation in order to secure fundamental labor reform. In July of 1978, Lula, aware that the increasingly soft-line military regime would soon open the party system to independent new entrants, wrote that a labor-based party in Brazil was just a matter of time (paraphrased in Secco 2011: 40). At the 1978 Oil Workers Congress in Salvador, Bahia, new unionists officially proposed the creation of the Workers Party for the first time (Secco 2011: 267). In January of 1979, new unions approved the creation of the party, overriding a small minority of dissenting voices (Keck 1992: 67). The push to create the Workers Party began, in earnest, in Following the new unions second general strike in March 1979, the newly inaugurated Figueiredo government ( ), instead of making policy concessions, 26 harshly repressed participating unions across the country, from Rio Grande do Sul to São Paulo to Minas Gerais. 27 Police arrested a large number of autênticos and imprisoned the new unionism s three principal leaders, 28 Lula da Silva, Olívio Dutra, and Jacó Bittar, for weeks. 29 Anti-labor repression galvanized the PT leadership and base. 30 The new union leadership formulated a statement of principles (Carta de Princípios) and distributed approximately 200,000 copies to attendees of the 1979 May Day rallies, held in most of the country s major population centers (Bahia, Ceará, Minas Gerais, Rio de Janeiro, Rio Grande do Sul, São Paulo) (Keck 1992: 67). By late 1979, a national, big-tent opposition movement had taken shape, which spanned 26 On his inauguration in March 1979, Figueiredo had reaffirmed the government s commitment to liberalization (von Mettenheim 1995: 106). In later months, he would make good on this promise. 27 Interview with Olívio Dutra in de Moraes and Fortes (eds., 2008: 122). 28 Sluyter-Beltrão (2010: 2). 29 Lula led the metalworkers union of São Bernardo do Campo, Bittar the oil workers union of Paulínia and Campinas, São Paulo, and Dutra the bank workers union of Porto Alegre, Rio Grande do Sul. 30 According to one leading new unionist and PT founder, the repression under Figueiredo made the new unions realize that they needed a political party; that strikes were not enough (interview with Bom). 78

87 elites in the political, cultural, and economic spheres and a host of interlinked civil society actors, including student movements, neighborhood associations, the Catholic left, Marxist parties and organizations, and at the forefront new unions (Skidmore 1988; Moreira Alves 1988). Despite their many internal differences, these groups united around the demand for democracy in order to force concessions from an increasingly soft-line military regime. 31 As the opposition movement reached peak strength, Brazil s controlled transition to democracy accelerated. In late 1979, amid social unrest and an economic slowdown, the Figueiredo government granted amnesty to all Brazilians still in exile and scheduled direct gubernatorial and mayoral elections for 1982, to be held concurrently with the congressional elections that had continued, with periodic disruptions and institutional manipulations, since the 1964 coup. In a new Organic Law of Political Parties, the government liberalized Brazil s party system, dissolving ARENA and the MDB, and allowing the formation of new parties for the first time since The Organic Law, which took effect several months after the May Day rallies, imposed steep electoral and organizational barriers to entry for new parties. In crafting the law, the Figueiredo government had two objectives: first, to divide the opposition and thus [stem] the rising tide of the MDB (Keck 1992: 87); second, to favor new parties built on the outgoing party establishment, while preventing the entry of outside political forces. The law stipulated that new parties with one-tenth of the national congress s current members (Chamber plus Senate) could obtain registry immediately. The rest, within twelve months of the law s passage, had to hold local party conventions (and thus to establish local offices) in twenty percent of the municipalities of at least nine states. To establish a local office, new parties had to incorporate, as members, two percent of the electorate in the smallest municipalities (i.e., those with fewer than 1000 voters) and Differences among these groups were minimized in the name of opposition unity (Keck 1992: 42). 79

88 members in the largest municipalities (i.e., those with more than 500,000 voters). 32 Parties would lose their registry if, in any future congressional election, they did not receive three percent of the national vote as well as five percent of the vote in nine states. Moreover, the law prohibited parties from mounting municipal campaigns in municipalities without a party Diretório. All of these provisions favored new parties built on old and outgoing established parties, as established parties could transfer politicians and infrastructure to their successors. The voto vinculado provision of the Organic Law stipulated that voters, on pain of ballot annulment, had to vote for the same party s candidates in every state or national race on the ticket. Thus, parties, in order to compete in a given state election, for example, had to field a gubernatorial candidate. This would force smaller opposition parties, in 1982, to place otherwise electable elites in gubernatorial races that they stood no chance of winning. Excluding the short-lived Partido Popular (PP), 33 five Brazilian parties obtained registry in advance of the 1982 congressional elections. All save the PT met the electoral or organizational requirements by inheriting legislators, members, and infrastructure from currently or previously established parties (Power 2000). Two of the four, the Social Democratic Party (PSD, est. 1979) and Brazilian Democracy Movement Party (PMDB, est. 1981), succeeded the ARENA and MDB organizations, respectively. The Brazilian Labor Party (PTB), to which deposed ex-president João Goulart had belonged, also resurfaced and split in two. Former PTB leader, Leonel Brizola, and a group of PTB elite allies and associated activist/patronage networks secured registry as the Democratic Labor Party (PDT) in Under the symbolic leadership of Ivete Vargas (niece of Getúlio), separate PTB networks with strong rural clientelistic linkages retained the PTB label and became registered 32 In addition, the law stipulated that in order to field state and local candidates, parties had to have offices in the relevant state or municipality. 33 The PP merged into the PMDB quickly after securing registry. 80

89 in Unlike the above parties, the PT lacked a partisan predecessor and thus could not count on a ready-made territorial network of local elites, activists, and branches. Most Marxist and Social Democratic leaders in Brazil including Fernando Henrique Cardoso and other leaders of the MDB s Tendência Popular argued against the creation of the PT, charging the party with dividing Brazil s progressive opposition and, in doing so, making the perfect the enemy of the good. Building the early PT organization From inception, the PT developed in adversity. Party founders lacked access to public and private resources and faced a closed, hostile, and highly influential mass media establishment. Paradoxically, these conditions created a powerful electoral incentive for early PT leaders to invest in mass party structures. The early PT stood as an exception in the Brazilian party system, putting a premium on grassroots organization-building and treating physical territory as the key arena of electoral politics (Meneguello 1989). Within a decade, the PT had built a powerful grassroots organization, with a dense infrastructural presence in several regions of the national territory and a large, committed base of card-carrying members. Low access to resources With a new and untested party label, a radical and adversarial profile, and the nationally established, center-left MDB on its right flank (Kitschelt 1994), the PT, on creation, attracted very few established left-wing political elites. Brazil s three largest Marxist parties, the PCB 34 Of these four, the PDS and PMDB remained serious national contenders in the long term and also produced offshoots the PFL/DEM (est. 1985/2007) and PSDB (est. 1988), respectively that became institutionalized as major national parties. 81

90 (est. 1922), PC do B (est. 1962), and MR-8 (est. 1964), immediately jumped on the PMDB bandwagon. In total, ten elected officeholders, all progressive MDB defectors, joined the PT in All ten held legislative posts at the federal or state level: one in the Senate, 35 four in the Chamber of Deputies, 36 and six in the São Paulo state assembly (Keck 1992: 82, 95). 37 Although the PT s MDB defectors provided valuable office space, telecommunications equipment, and administrative staff, 38 these organizational resources constituted a tiny fraction of what PT leaders would need in order to secure registry, much less to build a national party. In a country with over twenty states and over 4000 municipalities, the PT did not hold a single executive post in Strikingly, the PT continued with minimal state resources for nearly a decade. Prior to 1988, the PT elected only three mayors, two in small cities (Diadêma in 1982, Santa Quitéria do Maranhão in 1982) and one in the larger Fortaleza, Ceará (1985). Only in 1988 did the PT win a major subnational executive position, the mayoralty of São Paulo. In the same year, the PT won thirty additional mayoralties two in state capitals (Porto Alegre, RS; Vitória, ES) 39 and, for the first time, crossed the ten-percent threshold in Chamber seat share (10.2). In 1992, the PT won two small governorships, in the Federal District of Brasília and the state of Espírito Santo. The PT did not win the presidency or a major governorship (i.e., in São Paulo, Minas Gerais, Bahia, or Rio de Janeiro) before Adhemar Santillo of Goiás. 36 Edson Khair of Rio de Janiero, Antônio Carlos de Oliveira of Mato Grosso, Henrique Santillo of Goiás, Airton Soares of São Paulo 37 João Batista Breda, Sérgio dos Santos, Irma Passoni, Marco Aurélio Ribeiro, Geraldo Siqueira, Eduardo Suplicy. 38 According to Keck (1992), the early PT s MDB defectors provided critical infrastructural and logistical support during the period of party legalization (Keck 1992: 73). The benefits they brought with them were enormous and included meeting rooms, clerical help, free telephone service to other parts of Brazil, staff aides for research, mailings, and preparing other printed material. Most of the PT deputies dedicated as much as possible of this resource to party work (Keck 1992: 83). 39 RS: Rio Grande do Sul. ES: Espírito Santo. 82

91 The early PT also suffered from weak financing. From inception, the party depended for official revenue on four sources: (1) membership dues, (2) mandatory salary contributions (cotizações) from PT public officeholders, elected and unelected, (3) the federal party fund, with each party s financial allotment determined by its percentage of Chamber seats, and (4) volunteer contributions from individuals or corporations, whether outright donations or purchases of PT products such as shirts, pins, etc. From the early 1980s to mid-1990s, the majority of party funding came from salary contributions (Ribeiro 2010: 106). In March 1982, the PT mandated that petista public officials donate an extraordinary thirty percent of their salaries to the party (Secco 2011: 75), roughly ten times the percentage contributed, on average, by members of other Brazilian parties (Keck 1992: 133). During this time, the remainder of PT financing came primarily from the federal party fund (Keck 1992: 111; Secco 2011: 105-7), despite the PT s modest presence in the national legislature. The national Diretório formally required PT members to provide a small due (a cafezinho) to their municipal Diretório, but in practice, few ever did (Secco 2011: 105). The early PT explicitly eschewed ties to business and the economic elite, trumpeting itself as a partido sem patrões ( party without sponsors ). Corporate contributions would remain taboo (Ribeiro 2010: 109; Secco 2011: 107) until the mid-1990s and only become an explicit national objective in the 2000s. 40 Federal law prohibited formal linkages to, or donations from, unions. The ban on union donations deprived the PT of a revenue source available to many classic labor-based parties 40 According to one party founder and campaign strategist, the timing of the increase in corporate contributions resulted from several factors. First, in the mid-1990s, a large number of PT candidates had concluded that national electoral success would require adopting modo tradicional de fazer campanha. Second, many PT candidates had toned down their socialist rhetoric and lowered their negatives among the economic elite. Third, the PT, by the early 1990s, had won a number of important subnational positions, and in 1994, most observers believed that Lula would win the presidency. The business sector, according to the interviewee, wanted in on the PT in order to protect its interests. Governing experience also enabled PT politicians to meet businesspersons linkes to the public sector, whom party elites simply did not know during the 1980s (interview with Alli). 83

92 (e.g., Britain Labour Party, Sweden s Social Democratic Party). 41 Outside of formal party channels, new unions, which had deeper pockets than any other PT feeder groups or organizations, gave money to PT Diretórios, militantes, candidates, and cash-strapped local PT politicians, especially vereadores. 42 The new unionists directed such contributions primarily, but not exclusively, 43 to petistas with new union origins. Informal union contributions increased significantly after the creation of the CUT in 1982 (about which more below). One early party member recalled, for example, that the PT only began to use carros de som after the formation of the CUT. 44 Still, the PT suffered great penury until the late 1980s and remained financially weak until the second half of the 1990s. Annual party revenue ranged from 200,000 to 450,000 inflation-adjusted Brazilian reais between 1983 and From 1986 until the mid-1990s, annual revenue never exceeded 2.6 million reais (the figure in 1992) (Ribeiro 2010: 111). Consequently, countless PT fliers from the 1980s exhort militantes and supporters to donate money (in addition to campaign labor): Our party is poor, just like Brazilian workers. It lives off small donations from its members (filiados) and activists (militantes), who organize small parties and sell bônus. We don t use paid campaigners (Cabos Eleitorais) and run a decent, honest campaign. Campaign booklet for a slate of PT candidates in Rio de Janeiro state, We rely on the resources that we can create collectively. That s why each person s contribution is fundamental. Bulletin for Francisco Chico de Souza s state deputy campaign, In contrast to the other parties, which have the wealth from exploiting millions of workers to invest in their candidates campaigns...the PT and the Frente Brasil Popular [the PT s 41 Keck 1992: Interview with Secco. 43 Interview with Secco. 44 Interview with Kotscho. Ricardo Kotscho is an award-winning journalist and former Lula press secretary (2003-4). 45 Figures adjusted for 2010 inflation levels. 46 From CSBH archive. 47 From CSBH archive. 84

93 presidential electoral coalition in 1989] have as the only source of resources the voluntary contributions of its militantes and of the millions who will vote for Lula on November 15. Special financial bulletin released by the PT s state Diretório in Minas Gerais during the 1989 presidential campaign 48 In short, the formative-phase PT operated under severe resource constraints. Early party leaders could not provide financial or career inducements to activists save from officeholders salary contributions, jobs and resources controlled by a small number of governors, mayors, and legislators, and meager allotments from the federal party fund. In light of Brazil s vastness and population size, the patronage and finance held by early party elites could not begin to fuel a national electoral effort. The distribution of selective incentives thus played almost no role in the PT s development until the late 1980s, and a marginal role until the second half of the 1990s. Low access to media Whoever has television, radio, and newspapers will always be in power. Antônio Carlos Magalhães, Under military rule, television became the dominant medium in Brazil, particularly in the political news domain. Television ownership skyrocketed during the 1960s and 1970s, 50 growing by tens of millions. 51 During this period, Brazilians, by and large, did not use the radio to obtain news (Porto 2003: 292), 52 and only a tiny fraction of the population 48 From CSBH archive. 49 Bahian media mogul, three-time governor, and Minister of Communications under Sarney ( ) 50 The newspaper industry also expanded, although on a much smaller scale. Approximately one-hundred newspapers were founded during the last half-decade of military rule (Lawson 2002: 197). 51 Beyond Citizen Kane. 52 Broadly, Brazilians preferred television and print media as news sources, using radio for musical entertainment (Porto 2003: 292). Although some radio show hosts used their radio audiences to launch political careers (Porto 2003: 292), relatively few Brazilians consumed radio in order to obtain political news. Evidence suggests that radio has become a more important campaign vehicle in recent years (Boas n.d.). 85

94 regularly consumed newspapers or news magazines. 53 By the late 1970s, the vast majority of Brazilians learned about politics from television, 54 and ownership or indirect control of popular local/regional TV stations had become essential for political success. By law, television and radio networks could only own a handful of state and local stations. Vargas-era federal statutes gave the Brazilian president exclusividade not only in granting general broadcasting licenses to major networks but also in distributing the ownership rights of all state and local television and radio stations not owned by media corporations (i.e., the vast majority of the total). Moreover, the law allowed politicians and candidates to receive broadcasting concessions. Although politicians could not legally serve as station directors, they could, in their capacity as owners, appoint all station personnel (including directors) and thus dictate station content from behind the scenes. In 1993, Milton Tavares, a Brazilian lawyer, observed that in practice, politicians who owned local stations had long appointed supporters, friends, or family members to directorships, taking effective control of station content. 55 As mass media expanded in Brazil, control of broadcasting concessions became a major pa- 53 Brazilians might also have been more likely to view newspapers as politically biased. The director of Gallup Brazil stated in 1993 that television [in Brazil] has more credibility than print media for the simple reason that, for most viewers, images lend trustworthiness to news stories. The vast majority of the televisionwatching population, he stated, do not see a political position...in the different TV channels and, in general, think that what they see on TV is true (Carlos Matheus, Director of Gallup Brazil, quoted in documentary, Beyond Citizen Kane). Explaining public skepticism toward newspapers, one veteran reporter and PT member emphasized in an interview that Brazilian newspapers, in contrast to television programs, have never contained separate editorial sections and thus have always incorporated political biases quite perceptibly, in his view into their supposedly objective presentations of the news (interview with Kotscho). A petista functionary offered a different perspective, stating that Brazilian media, print and television alike, are more camouflaged than in the US. Citing Folha and Veja as examples, she claimed that Brazilians, even educated ones, wrongly believe that media sources, in general, are objective (interview no Newspapers and news weeklies affected politics mainly through elite channels, helping set elite agendas (Porto 2003: 291) but only contributing marginally to the mass dissemination of news. Television exerted vastly more influence on public opinion. 55 Interview in Beyond Citizen Kane. 86

95 tronage cow and source of power for the president. Presidents generally awarded television and radio concessions either to established allies (often politicians) or to local politicians or media outfits in exchange for political support. Rómulo Furtado, Secretary General of the Ministry of Communications from 1974 to 1990, summarized the presidential perspective: If I were president, I would never give a license or concession to an enemy. 56 Porto (2003: 293-4) summarizes that, beginning in the 1970s, the consolidation of urbanisation and of the cultural industry...originated a new kind of relationship between the state and local oligarchies. In exchange for allying with the president, local and regional political bosses received broadcasting concessions and began controlling the electorate...by using the local media, creating the new phenomenon of electronic coronelismo, as opposed to traditional coronelismo. 57 In order to maximize ratings, local and regional station owners had to rebroadcast popular national programs. By the time of the PT s formation, the Sistema Globo de Televisão, or TV Globo (est. 1962), held a national television monopoly. TV Globo owed its success to the military regime. Roberto Marinho, the CEO of the Globo conglomerate, had openly supported the 1964 military coup. 58 As a reward, Brazil s military governments had 56 Quoted in Beyond Citizen Kane. 57 Nunes Leal (1949) famously coined the term coronelismo to describe the clientelistic system that dominated rural politics in mid-twentieth century Brazil: one in which landowners secured the political allegiance of their rural workers largely through the distribution of individualized material benefits (e.g., credit, cash) and directed them to vote for particular politicians, who protected the landowners interests (e.g., by blocking land reform). A large number of scholars now use the term coronelismo eletrônico to capture powerful elites control of local and regional broadcast media and hence of their own political futures, given the increasing centrality of broadcast media to contemporary electoral politics. These include Stadnik (1991); Motter (1994); Costa and Brener (1997); Bayma (2001); Capparelli and dos Santos (2002); Porto (2003); dos Santos and Capparelli (2005); dos Santos (2006); Lima and Lopes (2007); Pieranti (2008); Stevanim and dos Santos (2011); Góes (2012); Boas (n.d.). 58 In the lead-up to the April 1964 coup often termed the revolução ( revolution ), especially by supporters the CEO of the Globo conglomerate, Roberto Marinho, openly sided with the military, despite having previously supported President Goulart. In an early 1990s interview, Armando Falcão, Brazilian Minister of Justice from 1974 to 1979, described Marinho as a revolucionário da primeira hora ( revolutionary in the first instance ) (from Beyond Citizen Kane. 87

96 nurtured TV Globo in its infancy 59 and, throughout the 1960s and 1970s, given TV Globo preferential treatment in the allocation of broadcasting spectra, federal subsidies and the non-enforcement of regulations (Lawson 2002: 184). The military had also made it impossible for other networks to compete commercially with Globo and been careful to avert the formation of any [competitors] (Amaral and Guimarães 1994: 33). As noted earlier, the Médici government canceled the broadcasting license of the Excélsior network in 1970, and months after taking office in March 1980, the Figueiredo government (1980-5) did the same to Rede Tupi (est. 1950), previously Brazil s largest television network and a Globo competitor in the make-or-break areas of telenovelas and news. 60 The only television networks besides TV Globo that thrived under military rule did so by assuming niches outside the TV Globo-dominated domains of telenovelas, sitcoms, and news. The seminational Bandeirantes (est. 1967) functioned as a specialized channel of sports, movies, direct sales, and packaged programming. The younger Sistema Brasileiro de Televisão (est. 1981) cautiously [avoided] any kind of competition with Globo and consolidated as a distant second place during the 1980s (Amaral and Guimarães 1994: 33). By the mid-1970s, TV Globo s flagship, prime-time news program, Jornal Nacional (est. 1969), commanded the highest nightly news ratings in Brazil by a wide margin. The network s news monopoly increased considerably during the second half of the 1970s due to the 59 TV Globo arose in 1962 two years before the military coup from a controversial agreement between the Globo media conglomerate (est. 1925) and the US-based Time-Life. In exchange for a share of future profits, Time-Life provided TV Globo with a massive infusion of start-up capital and thus an overwhelming advantage over competitors. Globo received six-million USD in start-up capital from Time-Life, compared to 300,000 USD for the Tupi network, then Brazil s largest television network. When the military seized power, TV Globo, trammeled by lawsuits resulting from the Time-Life deal, still did not possess legal status. In an explicit gesture of favoritism, the Castelo Branco government canceled pending legal investigations of TV Globo and awarded the network a television license in Rede Tupi produced Brazil s first major telenovela, O Direito de Nascer, in the mid-1960s. 88

97 placement of popular, Globo-produced telenovelas alongside Jornal Nacional in the primetime slot. Given the popularity of TV Globo programming, especially during prime time, most state and local stations, in order to succeed, had to obtain rebroadcasting rights from Globo. 61 Not only did TV Globo enjoy hegemony in the all-important television sector, the Globo conglomerate stood alone in holding top positions across hard news, television, radio, and print media. Almost all hard news came from the Globo News Agency, Brazil s largest, 62 and two São-Paulo-based newspapers, 63 the Estado de São Paulo and Folha de São Paulo. 64 The Sistema Globo de Rádio commanded the second largest radio audience (after Bandeirantes) and O Globo the third largest newspaper readership (after Folha and Estado). 65 Moreover, the non-globo news outlets, like Globo, had conservative owners and, conse- 61 By the 1970s, the path to political dominance at the state level involved two steps: first, offering political support in exchange for TV and radio licenses from the federal government, and second, securing the legal right to rebroadcast TV Globo content, since it [was] the dominant network (Porto 2003: 295). According to Amaral and Guimarães (1994), [i]f local stations [did] not partner with Globo, they [would] command a much smaller audience and have far less influence. [Joining] the weaker second- or third-ranked national systems...[was] a cut above standing alone, while autonomy...[was] synonymous with high costs, marginalization, and exclusion from the market (32). 62 Amaral and Guimarães 1994: Unlike in the United States, news weeklies (especially Veja) have also played a significant role in newsgathering (Porto 2003: 291). 64 Interview with Kotscho. 65 By the early 1990s, Bandeirantes and Globo (Sistema Globo de Rádio) owned roughly three quarters of Brazil s national AM/FM radio channels (Amaral and Guimarães 1994: 30). Bandeirantes owned 92 channels, the Globo network fifty. The other major radio networks Jornal do Brasil, Brazilian South, and Manchete owned 25, 17, and five, respectively. By 2000 (NEED OLDER DATA), Folha, Estado, and Globo s national newspaper (O Globo), boasted daily circulations of 472,000, 367,000, and 335,000, respectively. Globo did not enter the news weekly market, traditionally dominated by the conservativeleaning magazine Veja (est. 1968), until With Época, Globo quickly attained a top position. As of 2003, three news weeklies dominated the market: Veja (owned by the Grupo Abril, est. 1968), Globo s Época, and Isto É (owned by Editora Três, est. 1976), with readerships of 1,152,000, 498,000, and 381,000, respectively (Porto 2003: 291). 89

98 quently, 66 tended to leaned conservative in their news coverage. 67 These included SBT, Estado, Veja, and even Folha on some accounts. Ricardo Kotscho generalized, for example, generalized that a handful of conservative, traditional families have controlled news production in Brazil since before the military era, citing the Marinho family (Globo), the Mesquita family (Estado), the Civita family (Veja), and the Frias family (Folha). 68 The founder of SBT (est. 1981), TV host Sílvio Santos, openly espoused conservative views. 69 In summary, when the PT materialized on the political scene and initially developed at the end of the military era, TV Globo held a monopoly in television news, while a handful of predominantly conservative private media corporations, led by Globo, dominated the broader news market. In most Brazilian states, the leading television channel featured Globo news programming, including the prime-time Jornal Nacional and telenovelas, while the leading newspaper obtained hard news from the Globo News Agency. Very often, the same individual or group owned the leading television station, radio station, and newspaper (Amaral and Guimarães 1994: 30). Under these conditions, neutral or anti-establishment independent media might flourish in select, lightly monitored locales, but never at the national or state level. 66 Media observers, including within the PT, have often noted that that the conservatism of media higher-ups (the cúpula) generally carried the day and often created tension with rank-and-file journalists (the base) (interview with Kotscho). In São Paulo, [the media] are family dynasties, and in the rest of the states, the politics of the news business owners goes (faz-se). From interview entitled Meios de comunicação, balanço e compromisso in PT newspaper (CSBH archive). A 1986 campaign newspaper for PT candidates José Dirceu and Jorge Baptista, for example, states that [d]espite journalists resistance, the owners of the media manipulate information in accordance with their interests and only divulge the facts and version that serve the interests of the dominant classes (from CSBH archive). See also Meios de comunicação, balanço e compromisso, late 1980s PT newspaper (pp. 2-3), from CSBH archive. 67 One PT founder from São Paulo reserved special criticism for Globo and other broadcasters, on the ground that they, in contrast to newspapers and magazines, depend on public concessions and thus owe neutrality to the public (interview with Donato). 68 Interview with Kotscho. 69 In the late 1980s, Santos would seek the presidential nomination of Brazil s most right-wing party, the PFL, before aborting his candidacy. 90

99 Amid the opposition ferment of the late 1970s, TV Globo functioned as the military s unofficial mouthpiece. TV Globo systematically emphasized and occasionally exaggerated the military s accomplishments, especially on the economic front. Lawson (2002) summarizes this aspect of the implicit understanding between Globo and the military:...[the] Globo network played an important role in supporting the military s ideological objectives by advancing an image of Brazil as an economic powerhouse... In return, the network benefited from repeated allocations of broadcasting spectra and from the general growth of a heavily subsidized television market (Lawson 2002, p. 184) Equally, TV Globo omitted, distorted, and manipulated developments in order to weaken and defame the political opposition. In some cases, TV Globo s censorship extended beyond military dictates. For example, while the military forbade any airing of music by popular artist and regime opponent, Chico Buarque, TV Globo forbade mention of Chico Buarque s name altogether. 70 During the new unionists 1980 general strike, TV Globo drastically underreported the number of striking workers, misrepresented union demands, and aired statements from the industrialist, Mário Garnero, while cutting statements by Lula from the broadcast. 71 In a late 1980s interview on media, Jorge Baptista, journalist and PT leader, remarked that in a country of our dimensions, a political fact only becomes a POLITICAL FACT (emphasis in original) when it manages to pass through the meios de comunicação de massas In the PT s first high-profile attempt at municipal governance (in Diadêma), the party did not enjoy a sympathetic national press; there was extensive reporting of intraparty conflict and of the PT s mistakes in Diadêma, and very little coverage of successful efforts (199; for details, see 210). Remarkably, from early 1983 to early 1984, TV Globo did not cover 70 See Beyond Citizen Kane. 71 Beyond Citizen Kane. 72 Interview entitled Meios de comunicação, balanço e compromisso in PT newspaper (date unknown). 91

100 the Diretas Já! movement, which the PT led at the grassroots level. 73 Only in the final month of mobilization (April 1984) did Globo provide coverage, caving in due to ratings competition from other media outlets and mounting pressure from the public and from sectors of influential journalist unions, 74 including Globo s own. Moreover, before 1985, Brazilian parties possessed virtually no media platform of their own. Federal law prohibited parties from purchasing television time, and the Lei Falcão ( ) placed major restrictions on public radio and television advertisements, only permitting parties to display static party logos and candidate pictures/blurbs. In a context of media neutrality, the Lei Falção and prohibitions on the purchase of media time might have helped level the playing field, and thus benefited the early PT, by preventing the richest parties from pouring millions of dollars into slick, effective campaign ads. Given the conservatism of network coverage, however, these laws denied the PT the only opportunity it might have enjoyed to control a mass-consumed set of party messages and images. 75 In this way, the law magnified the political influence of Brazil s Globo-dominated mass media establishment. Given the legal constraints on public political advertisements and the legal prohibition of private television and radio ads, Brazil s dominant media networks, despite their dependence on public concessions, effectively filled the void, pushing a broadly conservative agenda and sometimes broadcasting open propaganda, both positive and negative, in order to influence electoral outcomes. The PT and media post-transition Relative to other media, television grew in political influence after the transition to civilian million individuals participated in the movement s final demonstration, on April 16, In a campaign bulletin from , a group of journalists running for union positions (and presumably affiliated with the PT) decried Brazilian media for their imbalanced coverage of the Diretas Já movement. The critique focused on the Globo network (bulletin entitled Jornalistas de oposição: Boletim de Campanha Chapa 2 (date unknown). 75 Interview with de Filippi. 92

101 rule. By , almost ninety percent of Brazilians obtained political information from television, either primarily or exclusively (Porto 2003: 291). Meanwhile, print media penetration remained weak. As late as 2000, only four daily newspapers were sold per one hundred citizens. In the early 2000s, only two thirds of Brazilians could read beyond a basic level, while only a third boasted more than a secondary education. Brazil continues to lag behind most of the world, including most middle-income Latin American countries (e.g., Argentina, Chile), in newspaper readership. 76 Given the outgoing military s relatively high level of support during the transition, especially from the economic elite and middle class, Brazil experienced a slow, incomplete transition to democracy, which the military largely imposed (Karl 1990). Relative to more discredited outgoing militaries in the region (e.g., in Argentina), the Brazilian military retained considerable powers post-transition (Stepan 1988; Valenzuela 1992). 77 Some scholars of the posttransition period have focused on the persistence of traditional politics (Hagopian 1996) and the military s efforts, through legal measures, to impose officialist successor parties and prevent the formation of external, anti-establishment parties like the PT (Mainwaring 1999; Power 2000). Others have focused on more specific issues such as budget allowances, weak civilian oversight of the military, and military control of the police, the intelligence services, and state enterprises (Stepan 1988). Relatively few scholars of the Brazilian transition have highlighted authoritarian enclaves in the media domain. 78 Some of President Figueiredo s final actions in office presaged what would follow under the Sarney and Collor administrations. In the waning days of military rule, Figueiredo awarded a number of licenses to political allies, even altering the planned allocation of broadcast 76 Porto (2003): Scholars have referred to these powers as prerogatives (Stepan 1988), tutelary privileges and reserved domains (Valenzuela 1992), and authoritarian enclaves. 78 Although see Porto (2003, 2012), who identifies the Brazilian president s effective retention of broadcast media prerogatives as a major and ongoing impediment to full democracy. 93

102 frequencies to different cities so that space on the radio spectrum would be available where it was politically expedient (Boas n.d.: 2-3). According to petistas José Dirceu and Jorge Baptista, these concessions included multiple radio channels for federal deputies José and Marcinino Camargo, the latter linked to Maluf and the PDS. 79 Under the Sarney administration ( ), the distribution of local television and radio licenses on political grounds reached a new level (Porto 2012: 63). The president and his communications minister, Antônio Carlos Magalhães, extensively used broadcasting licenses as political currency (Porto 2012: 63; see also Motter 1994). Of the 1028 TV and radio concessões that Sarney dispensed during his term, he distributed 539 (over half of the total) in the final nine months of the Constituent Assembly. Of these, he awarded nearly one-hundred to assembly representatives who, in exchange, supported the continuation of Brazil s presidential system and a one-year extension of Sarney s term (Motter 1994; Porto 2003: 294; Boas n.d.). 80 Broadly, Sarney and Magalhães awarded coveted concessions to state-level allies and supporters (sometimes family members), 81 conditional on legislative support, ground campaign help, advantageous local and regional media coverage, and/or other political favors. Although Brazil s 1988 constitution transferred control of concessões to the Communications Ministry and mandated congressional approval for awarding/canceling concessions, separate provisos such as the extension of licenses to fifteen years, exemptions for rebroadcasting stations, and quorum and voting rules for congressional approval allowed the media estab- 79 Meios de comunicação, balanço e compromisso, late 1980s PT newspaper, from CSBH archive. 80 During the first three years of the new democracy, President Jose Sarney and Communications Minister Antonio Carlos Magalhaes both powerful political bosses with media holdings themselves authorized broadcasting concessions for 91 representatives to Brazils Constituent Assembly. Evidence suggests substantial back-room dealmaking: many concessions were granted during the most intense months of deliberation on the new constitution, and legislators who received licenses went on to vote overwhelmingly in favor of several key amendments supported by Sarney (Boas n.d.: 3). 81 President Sarney awarded sixteen new licenses to members of his own family (Porto 2012: 63-4). 94

103 lishment and the Communications Ministry and hence the president to retain effective control (Amaral and Guimarães 1994; Porto 2003). 82 Amaral and Guimarães (1994) summarize that [t]he Constitutional Assembly of 1988 frustrated the dreams of those who backed the democratization of the Brazilian system of radio and television broadcasting via constitutional reform... [T]he wording of the 1988 law cemented the old system... [T]he possibility of change by administrative means is clearly difficult, if not impossible (Amaral and Guimarães 1994: 34). Five years after the Constitution s passage, Armando Rollemberg, president of the national Journalists Union, could generalize that political favoritism remains the only criterion for awarding concessões. 83 Analysts estimate that, as of the early 1990s, two thirds of local television licenses fell under the direct or indirect control of politicians who had received concessions from the federal government. 84 Moreover, in an analysis of twenty-one Brazilian states, Amaral and Guimarães (1994) found that in nineteen, a single individual or group owned both the leading newspaper and television station. The identities and political allegiances of Brazil s local and regional mass media barons did not shift appreciably after the transition to civilian rule. Under the Nova República, just as the traditional coroneis of the military era retained their rural clientelistic voting blocs (Hagopian 1996), the electronic coroneis of the military era, through connections to the Sarney and Collor administrations, retained media assets and alliances and moved to estab- 82 Porto (2003) writes: Although the new constitution included measures that restricted the politically motivated use of TV and radio licences, electronic colonelism continues to be a key feature of the Brazilian political system... Nevertheless, the constitution also established quorum and voting rules in order for Congress to reject licences authorised by the Executive, making it very difficult for the Parliament to reject them. The changes also left out of Congress control the repetidoras, the relay stations that retransmit the broadcasting signals of the networks. This lapse allowed presidents to continue using licences as political currency, as did Cardoso in his successful struggle to get Congress approval of his right to run for reelection. 83 Quoted in Beyond Citizen Kane. 84 Beyond Citizen Kane. 95

104 lish dominant positions within local and regional media markets (Boas n.d.: 2). In 1993, Armando Rollemberg, president of Brazil s Journalists Union, summarized that [t]oday, all the political groups that benefited from the dictatorship and the Sarney government are the owners of the television and radio in our country. This is bad because these groups no longer represent majority public opinion. The country changed... [It] rejected the military dictatorship. But the military s supporters have control of the media in our country. 85 Porto (2003) summarizes that the New Republic was...a period in which the legacy of the authoritarian past was not only maintained in some areas, but even strengthened. One example was the expansion of electronic colonelism through Sarney s presidency (293). Equally, TV Globo s influence did not wane under civilian rule. Both Sarney and Collor had personal ties to Globo that predated and fueled their initial political ascents. 86 During the military period, the Sarney and Collor families controlled media empires in Maranhão and Alagoas, respectively, and both Sarney and Collor, before rising to regional political prominence, had secured the exclusive right to rebroadcast TV Globo content in their home states. 87 Sarney and Collor thus owed their early political success, in large measure, to Globo. Moreover, upon assuming the presidency, both Sarney and Collor suffered from weaknesses low public support in Sarney s case, weak partisan support in Collor s case that increased their dependence on Globo and hence their susceptibility to Globo s influence. 88 The Sarney and Collor administrations manipulated regulatory enforcement to Globo s ad- 85 Beyond Citizen Kane. 86 Globo s subsidiary in Alagoas state, for example, played an important, perhaps decisive role in Collor s 1986 gubernatorial victory in Alagoas. 87 Sarney secured the rights before the 1985 transition, Collor in 1986 (in time for his 1986 gubernatorial campaign in Alagoas). 88 Argemiro Ferreiro, editor of the Rio-based newspaper, Tribuna da Imprensa ( ), has observed that the Globo network amassed immense power under Sarney because of the Sarney government s weakness and resulting susceptibility to Globo s influence (from Beyond Citizen Kane). Amaral and Guimarães (1994) summarize, along similar lines, that [t]he Globo network...imposed its structure as a national model on the civilian regimes that followed [military rule] (32; emphasis added). 96

105 vantage, actively contributed to the bankruptcy of NEC do Brasil and Globo s fire-sale purchase of the company, 89 and also gave Globo executives, particularly Marinho, direct and informal influence over relevant appointments and policy matters (Porto 2003, 2012). In the early 1990s, federal police arrested several individuals in Rio de Janeiro for watching a humorous, pirated television segment in which a narrator, dubbing over a video of Marinho speaking, showed viewers how to set up a pirated television transmitter. 90 In short, Roberto Marinho and Globo reached the apogee of their political power under Sarney and Collor (Porto 2012: 64). 91 Between 1988 and 1993, TV Globo commanded 78 percent of Brazil s television audience, with penetration in over 99 percent of the Brazilian territory, and with 99.9 percent of Brazilian televisions carrying the network. On the typical night, fifty-eight percent of Brazil s television audience, on average, tuned into TV Globo for the Jornal Nacional and telenovelas (CITATION NEEDED). 92 As of the early 2000s, television remained the most important medium in Brazil and TV Globo the dominant network, with an absolute majority of the national audience ratings (Porto 2003: 291). TV Globo s dominance increase[d] during prime time, and Jornal Nacional remained the main TV news bulletin in the country (Porto 2003: 291, 299). Consequently, politicians in the Nova República depended on access to TV Globo. Federal deputy, Paulo Ramos, observed in 1993 that for legislative candidates, appearances on Globo which Roberto Marinho 89 NEC do Brasil was a Globo competitor and subsidiary of Japan s powerful, multinational NEC. 90 Beyond Citizen Kane. 91 By the early 1990s, Marinho had become one of three Brazilian billionaires and, according to many (including Chico Buarque), the most influential civilian in the country (Beyond Citizen Kane). 92 TV Globo s popularity also stemmed from the beloved daytime show, Xou da Xuxa ( Xuxa Show ), which aired from 1986 to 92). During prime-time hours, twenty-four percent of the national television audience watched SBT (Porto 2003: 291). Along with SBT and Bandeirantes, TV Globo controlled 93 percent of the television stations in Brazil and commanded 98 percent of Brazil s television audience (Amaral and Guimarães 1994: 28). During the mid- to late 1990s, Globo received far more television advertising revenue than any other network, with approximately two-and-a-half times that of SBT, its nearest competitor (Porto 2003: 291). 97

106 occasionally vetoed often made the difference between victory and defeat. 93 In implicit exchange for government favoritism, Globo awarded rebroadcasting contracts to allies of Sarney and Collor. In 1986, after Sarney s communications minister, Antônio Carlos Magalhães (ACM), helped engineer the bankruptcy of Globo competitor, NEC do Brasil, TV Globo broke an 18-year contract with TV Aratu and awarded the exclusive rebroadcasting rights to ACM s own TV Bahia (Porto 2003: 295). More importantly, Globo and other networks continued to broadcast conservative news coverage and even open propaganda, despite the passage of important post-transition reforms, which gave parties legal tools to respond to such abuses. TV Globo adopted a frequently laudatory, infrequently critical posture toward the Sarney government, despite allegations of rampant corruption and cronyism and the deteriorating macroeconomic conditions of the late 1980s. In a 1986 interview, Paulo Frateschi, joint organizer of Apeosep, the São Paulo state teachers union, observed that Sarney s stabilization package in the late 1980s passed with TV Globo, Manchete, Folha, Estado, everyone there, displaying the plan as if it were the great salvation. 94 Only when commercial interests dictated a critical approach (e.g., during the movement to impeach Collor) did TV Globo provide more access to the opposition. TV Globo s most flagrant and egregious breach of journalistic neutrality took place during the 1989 presidential election campaign, which pitted Collor against the PT s Lula da Silva. In the campaign, Lula [had been] vilified day and night by the press, and not just Globo (Secco 2011: 141). The management of Estado de São Paulo, for example, imposed a pro-collor line not just a pro-collor editorial position on the paper s rank-and-file journalists. 95 It was TV Globo s unbalanced coverage of the Lula/Collor debate, however, that likely tipped the 93 Beyond Citizen Kane. 94 Published interview entitled, A Dívida É Fabricada (1986), from CSBH archive. 95 Well-known reporter Augusto Nunes stated this publicly in 1988 (interview with Kotscho). 98

107 polls in Collor s favor. 96 In its prime time recap of the debate, the Globo s Jornal Nacional gave seventy seconds more air time to Collor and, more importantly, heavily edited debate segments so as to increase Collor s favorables and portray Lula as radical and dangerous. Vianey Pinheiro, producer of TV Globo São Paulo at the time, called the Globo s prime time coverage of the debate which he participated in editing a publicity special for Collor. President of the national Journalists Union, Armando Rollemberg, called the coverage a shameless manipulation and flagrant breach of journalistic ethics. 97 During the 1989 campaign, the PT produced a television parody of Rede Globo called Rede Povo, displaying, with great irony and creativity, facts...omitted by the mass media and alternative perspectives on the big national issues (de Moraes and Fortes, eds., 2008: 282). It should also be noted that, after the 1985 revocation of the Lei Falção, access to public television remained highly unequal, a point frequently stressed by PT candidates after the transition to civilian rule. A representative 1986 campaign flier for Clara Ant, PT candidate for state deputy in São Paulo, decried the anti-democratic division of free time on TV as one of many maneuvers against the PT. 98 In many cases, publicly funded television programs remained under the control of conservative individuals and groups. In a late 1980s interview, PT leader José Dirceu observed that with regard to curating personnel, public educational TV in Brazil and São Paulo s influential RTC (Rádio e Televisão Cultura) in particular remained in the times of the military dictatorship. 99 In short, following the 1985 transition from military to civilian rule, the media environment 96 In a conversation with the author, a São Paulo-based journalist stated that, in his view, Roberto Marinho s involvement in the 1989 election decisively affected the outcome (interview no. 6). 97 The content and quotes in this paragraph come from Beyond Citizen Kane, a 1993 BBC documentary censored in Brazil (first fully and then outside universities) throughout the 1990s. 98 From the CSBH archive. 99 Meios de comunicação, balanço e compromisso, late 1980s PT newspaper (date unknown: p. 3), from CSBH archive. 99

108 did not improve for the PT. As late as 2002, Brazilian mass media remained profoundly biased toward candidates of the [r]ight (Lawson 2002: 200). PT leaders during the early stages of the Nova República constantly decried the politics of Brazilian media and pressed for media reform: Many petistas who ran for legislative and Constituent Assembly seats in listed the democratization of media as a central platform point (e.g., Plínio Arruda Sampaio, Geraldo Siqueira). Summarizing a widespread view within the PT, Jorge Baptista, journalist and petista, commented on possible media reforms in an undated, Sarney-era PT newspaper interview: One [way forward], mainly in relation to broadcast media, which are the most powerful, is through the control of concessions today a privilege of the president...used to benefit the most well-connected (chegados), the closest to home (os mais intimos da casa). 100 A 1992 PT election campaign manual stated: In Brazil, elections have long been characterized by radio and television abuses. In numerous elections, we have witnessed numerous instances of electoral propaganda in radio and television programs, which end up directly interfering in the electoral process... We know that...the stations (emissoras) will highlight some candidates. What we cannot allow is for the stations, on the pretext of providing news, to do open propaganda for their favored candidates and against their opponents. 101 In a 1993 interview, Lula identified the hegemony of Globo as a central impediment to full democratization in Brazil: Democracy presupposes freedom of communication, speech... There won t be democracy without the democratization of mass media... If you have one outlet that every day talks to sixty-, seventy-million people, and the control of the messages falls to a team taking ideological orders from one man [Roberto Marinho]... one is deprived of any possibility of democracy. 102 * * * In summary, under military rule and the early Nova República, the PT confronted an influential, reactionary, Globo-dominated mass media establishment, at turns closed and hostile to the left opposition. Media hostility impeded the PT s electoral progress, and low media access constrained party elites. Some analysts of Brazilian media have stressed the personal convictions of individual media owners, particularly Roberto Marinho, in explaining the conservatism of mass media 100 Meios de comunicação, balanço e compromisso, late 1980s PT newspaper (pp. 2-3), from CSBH archive. 101 Instruções sobre propaganda eleitoral for Eleições 92 (date unknown). 102 Beyond Citizen Kane. 100

109 in Brazil. 103 This argument smacks of excessive voluntarism. The longstanding structure of state/media relations in twentieth-century Brazil not the chance preferences of media moguls virtually ensured conservative dominance of Brazil s airwaves and newspapers during the PT s formation. Media corporations primarily sought profit, and in order to run profitable media enterprises, they depended on, or at least benefited greatly from, the good favor of the political establishment. The Globo conglomerate, for example, has long prioritized commercial interests over any particular ideological agenda. Throughout its history, Globo has cultivated relationships with the governing elite, whatever the corresponding ideology. Rádio Globo received its initial concessions from Presidents Kubitschek and Goulart, to whom Roberto Marinho enjoyed regular personal access. The military regime nurtured and favored TV Globo. After military rule, Globo enjoyed a cozy, mutually advantageous relationship with Sarney and Collor (1990-2). When public opposition to Collor became overwhelming, the Globo network piled on for commercial reasons. Argemiro Ferreiro, editor of the Rio-based newspaper, Tribuna da Imprensa, summarized in the early 1990s that the hegemonic Globo [was] above all a commercial enterprise. Globo, he summarized, aligns itself with whomever is in power, possesses a governista vocation, and functions as a kind of official, or quasi-official, spokesperson for the powerful (os donos do poder). 104 In closing, it should be added that for the first decade and a half, the PT, by and large, did not pursue or make effective use of the scarce media opportunities that it did have. In the 1982 subnational elections, the PT s first major electoral test, [p]robably the least effective element in the PT s [São Paulo] campaign...was the use of the mass media... The party was 103 Amaral and Guimarães (1994); vanden Heuvel (1995); interview with Kotscho. Several scholars (e.g., Porto 2012) have observed that Roberto Marinho and other top Globo officials personally held conservative political views and used their influence to shape political developments. 104 Beyond Citizen Kane. 101

110 not particularly successful in linking electoral positions with topical issues of the kind likely to arouse journalistic interest. Few candidates succeeded in making use of those spaces that might have been available TV talk shows, for example (Keck 1992: 143). 105 Although the PT did become more media-friendly and -savvy over the course of the 1980s, 106 the party s relationship with media remained fraught. In , the PT boycotted the grande mídia. 107 Only in the mid-1990s did the PT end its war with mass media and begin to court all major networks actively, recognizing the importance of media not only for campaigns but also for governance. 108 The incentive for organization ORGANIZATION Full caption, large and in bold, under Lula headshot (PT campaign sticker, 1982) During the PT s formative years, the Brazilian political context dictated that any externally created party of the left must invest in organization in order to have a chance at success. Given the influential and reactionary mass media establishment, no party conceived outside state structures, media-savvy or not, could have acquired a national following through media appeals. Only a grassroots party like the PT could have succeeded. Selection effects thus help explain the PT s organizational strength. Yet the PT s investment in grassroots organization also stemmed from the explicit recognition among party founders a recognition powerfully evidenced in campaign materials, 105 The party had a great deal of trouble getting its message across in newspaper coverage of the campaign; news coverage tended to either report party statements flatly or to concentrate on the conflict between the PT and the PMDB (Keck 1992: 143). 106 Some petistas cite Eduardo Suplicy s 1985 campaign for the São Paulo prefeitura to illustrate the party s media learning. According to one party founder, Suplicy s segment during the horário eleitoral gratuito remains a PT model to this day (interview with Azevedo). 107 Interview with Kotscho. 108 Interview with Kotscho. 102

111 internal party communications, and member interviews that low access to resources and a hostile, closed media environment made organization fundamental. The previous sections have attempted to show that the early PT lacked access to state and private resources; that Brazil s top media networks ignored and defamed the PT; that the PT, like all Brazilian parties, could not purchase television or radio time; and that for free public advertisements, the PT, initially, had to use a limited, ineffective platform and, after the revocation of the Lei Falcão, still had to cram party and candidate advertisements into very short time slots. The current section demonstrates that these conditions affected PT founders organizational calculus. Materially strapped, off the airwaves, and in competition with politicians and parties that enjoyed considerable resources and media access, early PT leaders understood that the party s rise or fall would depend on the size, efficacy, and durability of its grassroots structures. Archival research revealed thousands and suggested the existence of hundreds of thousands of campaign materials and internal communications stressing (1) the PT s penury, (2) the mass media s manipulation of political news, boycott of the PT, and hostility to the PT, and (3) the opposition s great wealth and extreme media favoritism. Crucially, PT elites and activists also explicitly stressed that these difficulties rendered organization vital. Copious campaign pamphlets, activist manuals, internal analyses as well as interview excerpts and external analyses draw the connection between the PT s low access to media and resources and its reliance on volunteer activism and contributions: 1982 elections The PT was a partido sem patrões... Because of its very nature, it had financial difficulties. This...was overcome (superada) by the efforts of the miltantes, who donated money, who raised money...and who had a genuine identification with and commitment to the party... The militancia was the only thing I had in 1982, the volunteer militância [i.e., during his bid for federal deputy in São Paulo]. [Later in the interview] There was no space at all (espaço nenhum) in the small, medium, or large media. What is the media in Brazil? It has always been in the hands of the richest, most traditional families. Even local media. They didn t interview us, cover us, or inform people about us. But they did cover 103

112 other parties. There was simply no coverage of us. interview with Djalma Bom, new unionist and PT federal deputy (1982-6) [Barroso] depended exclusively (contava somente) on the sacrifice and effort of comrade groups in Vila Prudente, the ABC region, and other districts... [H]e didn t have resources bulletin from the Electoral Office of PT São Paulo municipal councilor, Cláudio Barroso 109 The PT doesn t have the support of the powerful. Our campaign depends on the strength and union of our supporters campaign paper entitled Why to vote for the PT, section entitled Participate in the PT campaign 110 For workers, a strong and organized political party is fundamental in order to win power... In these elections, our tool is the PT. PT campaign flier 111 Only the organization of us workers will guarantee that our plan is realized. Campaign booklet for Virginia de Angelis elections We re reaching the finish line of this electoral campaign. More than ever, economic power counts (se faz valer). The PT s campaign evidently depends on other powers: the energy and will of those who believe in this platform... THE DECISIVE MOMENT HAS ARRIVED. We re going to need all the enthusiasm, availability, creativity and organization to take the platform of the Eder candidacy to as many people as possible. Campaign bulletin for Eder Sader, federal deputy candidate from Rio de Janeiro The PT doesn t have the support of the rich and powerful for its electoral campaign. We need your participation in order to spread the propostas of the PT. We depend on you. Campaign flier for Eduardo Suplicy, SP gubernatorial candidate, and Hélio Bicudo and Jacó Bittar, SP senatorial candidates 113 Materials from Francisco Chico de Souza s 1986 state deputy campaign: All the economic power of the bourgeoisie is used to finance the millionaire campaigns of its candidates. Campaign bulletin A campaign flier emphasizes that the real world differs from what appears on television In contrast to other parties, we rely exclusively on the strength of our militantes (Diferentemente dos outros partidos, nós só contamos com a força dos nossos militantes) Letter to militantes, Nov. 5, From CSBH archive. 110 From CSBH archive. 111 From CSBH archive. 112 From CSBH archive. 113 From Edgar Leuenroth archive. 114 From CSBH archive. 104

113 In the electoral campaign we ll have to confront the economic power and electoral machines of the patrões, even the anti-democratic division of free propaganda time on TV (one of the many maneuvers against the PT). It will be with the collaboration and effort of the workers themselves that the PT carries its message to the population. Campaign flier for Clara Ant, São Paulo state deputy candidate Now the campaign begins in earnest. We re going to put our team on the field. With enthusiasm, with desire. We re going to confront economic power with creativity and intense work. Campaign bulletin for Olinto Alves Leite and Wallace Dellamagna Sant Ana, candidates for federal and state deputy in São Paulo, June Early campaigns broadly The PT had much less money than the other parties... The militância was everything. interview with Ricardo Azevedo, founding PT activist and intellectual My mandate was always very characterized by the grassroots (organicidade), exactly what I had built for ten years... I was always a public figure, but not very much of the mass media. My process for building a mandate was always very linked to the grassroots party and the grassroots union movement. Paulo Rocha, new unionist and five-time PT legislator from Pará, describing early campaigning experiences presidential election (Lula v. Collor) This is a campaign with scarce economic resources because the PT is a poor party that depends on the contributions of its members, us, the workers. That s why your support is fundamental your desire to work and publicize the PT and Lula. From PT newspaper for Lula s presidential bid 117 The PT s official campaign manual for the 1989 presidential election states that, in the first phase of the campaign, the entire party, each diretório municipal, núcleo, militânte or filiado of the PT...has the responsibility to carry to the streets the true and only hope of change, which is Lula s candidacy (1). The manual instructs militantes to contact local radios and newspapers and feed them campaign information on a daily basis, if possible. This scheme is fundamental for the PT and for the Frente, given that the mass media s boycott of [our] campaign is big (4). Apart from the local media strategy, the manual focuses entirely on mobilization, describing the third and final phase of the campaign as the moment of mass acts, rallies, agitações, and marches. 118 In a manual for PT activist leaders, the Comitê Popular Pró-Lula treats volunteer activism and mobilization as the core of the PT s 1989 campaign strategy. The manual emphasizes that the party s financial constraints for popular parties and campaigns...always the biggest problem necessitate the following measures: creative money-saving (e.g., using one s own typewriter to produce party leaflets); finding free or reduced-rate meeting places belonging to a companheiro; soliciting contributions from other militantes and supporters; hosting fundraising barbecues and parties; selling 115 Segment entitled, Vamos Fazer Uma Campanha Criativa. From CSBH archive. 116 Interview with Paulo Rocha in de Moraes and Fortes (eds., 2008): From the CSBH archive. 118 Page number of final quotation not recorded by author. From CSBH archive. 105

114 party paraphernalia; forming electronic brigades, making homemade videos, and showing them in homes, clubs, unions, communities, and even in public plazas ; skillful wall-painting in order not to create enemies ; and forming distribution groups to take printed materials door-to-door and to areas with high concentrations of people, including factories, fairs, public transportation stops, schools, busy streets, and avenues. 119 Mário Milani of Paraná s Coordenação de Propaganda distributed a manual to the state s municipal campaign committees. The manual stated: It is known that the printed media (newspaper, magazine, and outdoor) and electronic media (radio, television, and cinema) are in the hands of the bourgeois parties... The big propaganda agencies, the bankers, the multinationals, the agriculturalists, and big business are armed with dollars to make their ideas count... If we use creativity, we will be able to spread our proposals more efficiently than they, who command the mass media. The manual proposes specific strategies, encouraging militantes, for example, to put up pictures of fists with a large L (for Lula) and filling in the ula later. 120 The issue of campaign financing, in a party like ours, places on us the responsibility of linking the issue tightly to the party organization and the organization of popular support committees for Lula. We have to turn the financial question into an instrument of organization for our support base. This base cannot be composed solely of militantes for the parties in the Frente Brasil Popular, but must expand to the point of organizing a network of contributors, willing to build a left alternative in the country... It falls to the PT to motiviate the Frente and all the parties that compose it to embark, with boldness (ousadia), on a fund-raising drive in order to avoid being suffocated by the economic power of our adversaries. Collor, Brizola (PDT), Ulysses [Guimarães, PDT] and others have the machines of the bourgeoisie to sustain them: planes, million-dollar fundraisers, free time on TV, press coverage happen one after the other under our eyes. To contain this situation an unprecedented activist effort is necessary [emphasis not in the original]. Every Lula supporter must be a fund-raising post. Imagination must be placed in the service of seeking power. Dinners, parties, individual fund-raising, lists, bazaars, barbecues, feijoadas must multiply across the country, in order to create a true pro-lula current. PT presidential campaign bulletin, no. 9 of Organizacções Globo, and especially the Rede Globo de Televisão, give prominence to Collor s initiative against some of Alagoas s marajás [i.e., idle, well-paid bureaucrats]. The link between his family and Globo is old: his father funded a media empire in Alagoas... His brother Leopoldo was Globo s commercial director in São Paulo. The political bond between Collor and Roberto Marinho was consummated in July 1988, with lawyer Jorge Serpa acting as intermediary... [Marinho and Serpa] manufacture a name capable of deluding a segment of Brazilians...who can give the impression of rejecting traditional politics. For this contract job (empreitada), they have relied on the directors of powerful state companies like Osires Silva, ex-president of Embraer and Petrobrás, on owners of the big construction companies (Murilo Mendes, Sérgio Quintela), on leaders of the most reactionary business sectors (like Rui Barreto, ex-president of the Associação Comercial do Rio and Albano Franco, president of the Confederação Nacional da Indústria), on the ministers of the Army, Leônidas, and Communications, Antônio Carlos Magalhães. The manufactured candidate... [At the bottom of the paper] The Movimento Pró-Lula s finances do not receive help from the wealthy and the speculators: they depend on the solidarity of militantes, supporters (simpatizantes), and voters (eleitores). Raise money in all the activities you carry out, get contributions from friends, and send this suado e carinhoso dinheirinho [ kind, hard-earned small sum ] to the account Lula 89 PT, no , Ag. 0300x, São Paulo, Banco do Brasil PT newspaper entitled Brasil Urgente, Lula Presidente 119 From CSBH archive. 120 From CSBH archive. 121 From CSBH archive. 106

115 [B]y 1988, political commentators counted the [PT s] ability to call upon its activists for social and political campaigns as a political resource that almost made up for its lack of financial resources. Keck (1992: 238-9) Perpetually short on funds, the PT [in 1989] depended heavily on the commitment and energy of its militants Hunter (2010: 113) 1994 presidential election (Lula v. FHC) From Os 13 pontos do Rui Falcão: Manual de Campanha (the campaign manual prepared by PT president, Rui Falcão, for the 1994 presidential race): FHC is the candidate of Itamar, of the sharks in power under Collor who tried to stop his impeachment. For these candidates of the rich, money and colorful propaganda and the support of the TVs, radios, and newspapers won t be lacking. Against them, we will put on the field the enthusiasm (garra) of the working class, the strength of the union movement, the energy of popular organizations, the thirst for change from a nation that can no longer stand so much injustice and corruption... The large media networks and the press, repeating what they already did in 1989 to elect Collor, openly defended the government-supported candidate, Fernando the Second [i.e., FHC]... But starting now we have a small slot on radio and television, to show the reality of things and strike down accusations. Even more important: starting now, the famous and feared militância do PT, whom our enemies so criticize but observe with such envy, wakes up and enters the contest... In a campaign with few resources, like ours, each of our supporters homes needs to function as a kind of Residential Committee, with materials to distribute, propaganda out front, meetings to talk about the campaign and prepare propaganda activities The candidates of the rich and corrupt run their campaigns giving money to paid electoral workers (cabos eleitorais). We run campaigns with the voluntary sweat and dedication of our militantes municipal elections [Our] moral victory [in the 1996 municipal elections] came with the creativity and political will of the militância and our radical posture of confronting political and economic power with the petista mode of campaigning: politicization of the debate, social mobilization, a campaign of activism (campanha militante)... The election brought back rallies, marches, the PT star and flags, from Porto Alegre to Belém, from Campo Grande to Natal. Um Balanço Petista by José Dirceu (reflections on the 1996 elections) 123 Access to civil society and the capacity for organization 122 From CSBH archive. 123 From the Edgar Leuenroth archive. 107

116 Low access to resources and media gave the PT an incentive to build a strong party organization. Access to civil society in particular, to the new union movement, the grassroots Catholic left, and elements of the Marxist left gave party founders the capacity to succeed. With rare exceptions and slight variations, PT members and scholars of the PT s formation converge in identifying the new unionism, the Catholic Church, and Marxist left (often called the traditional left or organized left) as the three core feeder groups, or vertentes, of the PT. 124 Due to time and resource constraints, the early PT leadership could not make a significant material or logistical contribution to the formation of local organs and nuclei. Major national leaders visited local organs and civil society organizations to give speeches and provide encouragement and advice, 125 while the major Diretórios disseminated manuals for setting up and sustaining nuclei, Diretórios, and electoral committees. 126 But this was all. The party cúpula did little more than encourage local civil society leaders to build local nuclei and Diretórios from the ground up to assume party leadership roles, to draw volunteers and administrative resources from their own organizations and movements, and to pool participating individuals resources, labor, and creative energies. Keck (1992: 94) thus summarizes that, in order to maximize territorial expansion, early PT leaders [appealed] to leaders of already organized constituencies, mainly leaders of unions and grassroots movements, members of the CEBs, and members of the organized left. A 124 Sobre a construção partidária, Raul Pont, June 1987 (from Edgar Leuenroth archive). Also, interview with Sergio Alli. 125 Although the national leadership faced time and resource constraints, major PT leaders (including Lula) did play an important ground-level role by visiting established local party organs or local civil society organizations to encourage leaders to create organs. In Paraná in 1982, for example, Lula once visited twelve municipal Diretórios in a single week (Secco 2011: 59). Rural union leader Manoel de Conceição played a critical, ground-level role in the formation of the PT in Maranhão and Rio Grande do Norte (Secco 2011: 59). 126 A PT text from Belo Horizonte on electoral tactics, dated Christmas 1980, stresses the urgent need for the party to adopt a policy of quadro formation, of political consciousness-raising, so that in each municipality our companheiros can walk on their own two feet without feeling disoriented (from the CSBH archive). 108

117 PT founder and unsuccessful candidate for governor of Pernambuco in 1982 captured the early PT s reliance, for organization-building, on local civil society leaders: [The party] grew little by little... [W]e built those PT nuclei, beginning from our relationship with leaders in each place... The shift from boss (cacique) to collective must be a long, organic trek, in the union, in the cooperative, in the association, in the neighborhood, in the community, in the settlement, in the school, wherever there are people to organize (emphasis not in the original). 127 In short, the PT s first decade of organizational expansion occurred via diffusion (Panebianco 1988: ): in a decentralized and spontaneous manner, led by distinct party elites in different regions, without the presence of a strong center in charge of expansion (Ribeiro 2010: 251-2). In building its organization, the PT, in effect, created a new class of political elites. The PT cúpula called local civil society leaders and activists into the institutional sphere, and through internal promotion within the party organization, these leaders and activists ascended to major leadership positions and became party candidates. 129 A flier for petista Luiz Gushiken s 1986 federal deputy campaign stressed that [i]t s not enough just to elect our candidate. We seek to strengthen the PT, with the training of hundreds of petista leaders (Edgar Leuenroth archive). Keck (1992) observes that as of the early 1990s, a growing number of petistas were rising to leadership positions from within the party organization. The PT and unions The new union movement acted as the fulcrum of the PT s decentralized, grassroots, partybuilding endeavor. Throughout the PT s formative phase, but especially in the early 1980s, 127 Interview with Manoel de Conceição in de Moraes and Fortes (eds., 2008): According to Panebianco s original definition, territorial penetration, as distinct from territorial diffusion, occurs when the center...leads...the formation of local...party associations. Territorial diffusion occurs when...local elites construct party associations which are only later integrated into a national organization (Panebianco 1988, p. 50). 129 Author s conversation with Ribeiro,. 109

118 new union leaders became PT leaders, new union rank-and-file became PT rank-and-file, and new union offices and members homes became PT offices and meeting places. New unions provided the PT with more leaders, quadros, activists, and rank-and-file members than any other civil society movement or organization. In addition, new union subculture and institutions, including the base-level internal democracy that had distinguished new unions for years, carried over into the PT (Ribeiro 2010; Secco 2011). The PT s new union base expanded and evolved considerably during the PT s early development. In 1983, new unionists founded the Central Única dos Trabalhadores (CUT) in order to strengthen the new unionism. The autonomous labor movement had already begun to extend beyond its original bases (e.g., ABC paulista, Porto Alegre) to new regions and economic sectors. New unionists intended the CUT to accelerate this process. The CUT founding thus provoked the military s last gasp of hard-line retrenchment. The Figueiredo government struck at the heart of the just-formed confederation in 1983, intervening in an ABC metalworkers strike and arresting the top leadership, including Lula. Later in the same year, the CUT carried out its first general strike. Approximately two-million workers participated, suggesting that the social force of the new unionism had already increased (Secco 2011: 268). Over the next several years, the CUT expanded, and autonomous labor grew in stature. The CUT surpassed CONCLAT to become Brazil s largest union confederation. The growth of the new unionism largely accounted for the expansion of the PT base and infrastructure, and for the PT s increasingly national character, over the course of the 1980s. In part, autonomous labor grew through the incorporation of rural unions, thanks to the organizing efforts of the Catholic left s Pastorais da Terra. 130 Prior to the 1983 founding 130 A PT founder from Pará, for example, recalled the integral role that CEBs played in bringing rural unions into the CUT:...[W]ith the founding of the CUT-Pará, I came to lead its process of construction. In 1986, we won twenty-four rural unions, naturally, with much help from the CEBs, which were strong in Pará (Interview with Paulo Rocha in de Moraes and Fortes (eds., 2008): 149). 110

119 of the CUT, 131 urban new union leaders decided, somewhat contentiously, to include rural unions in the confederation. 132 Strikingly, more than a quarter (27.3 percent) of the founding CUT signatories were rural unions (Secco 2011). More centrally, the new unionism grew through the incorporation of public- and middle-sector unions. During the late 1970s, private-sector industrial unions, typified by Lula and the metalworkers, dominated the new unionism both numerically and in the popular imagination. For most of the 1980s, however, the paradigmatic labor struggles involved bank workers and teachers (Keck 1992: 193). Correspondingly, after Lula s long stint as the PT s first president (1980-7), two bank workers Olívio Dutra (1987-8) and Luiz Gushiken ( ) occupied the party s top internal position. As the autonomous labor movement became more middle-class, it became less confrontational, which spawned the creation, in 1986, of a leftist CUT faction, the CUT pela Base, in São Paulo (Secco 2011: 268). Although the moderation of autonomous labor led to similar changes in the PT s base, program, and rhetoric and the development of a softer, more inclusive external image (Keck 1992: 193), it also periodically created tension with the PT. Despite considerable overlap between the PT leadership and new unionist/cut leadership, on occasion the PT criticized the CUT for not taking more radical positions in favor of workers interests (Keck 1992: 193). During the general strike of 1987, for example, a number of top PT leaders publicly rebuked the CUT leadership for not exhorting rank-and-file unionists to participate. The above notwithstanding, the labor movement remained the PT s backbone. In contrast to CONCLAT, whose leaders political allegiances varied, the CUT remained basicamente petista (with minority participation from other groups) through the 1980s (Secco 2011: 68). 131 New unionism s umbrella confederation (about which more later). 132 Interview with Hamilton Pereira in de Moraes and Fortes (eds., 2008: 266). 111

120 Given the internal diversity of the new unionism, the profile of the PT s main feeder unions varied considerably across regions, within regions, and even within states. In the ABC paulista and other urban areas (e.g., Betim, MG), industrial new unions played the dominant role in building the early PT. In other, less industrial urban centers, middle-sector new unions often played the dominant role (e.g., bank workers in Porto Alegre, teachers in Rio de Janeiro and Belo Horizonte). 133 In the Northern region, rural unions, in conjunction with the Catholic Church, tended to lead the process. In a few places, the PT created new unions, not the other way around for example, in Maranhão, where the state s dominant PT tendência, the PT de Aço, under the leadership of Manoel de Conceição, recruited rural workers into the new union movement. 134 It should be emphasized that most Brazilian unions did not join the PT or the new union movement. Many Brazilian unions had conservative leaders who rejected new unionism in order to protect their own privileged relationships with the political establishment. A significant minority had left-wing leaders who, instead of supporting the new union movement and PT, supported or belonged to Marxist parties, above all the PCB. The PT and the Catholic left The Catholic left played a vital role in the organizational development of the PT especially in urban peripheries and rural areas. In the formulation of Frei Betto, 135 CEBs possessed autonomy and complementarity in relation to the PT (as to the new unionism previously). CEB members affinity with the PT largely stemmed from the characteristics and experiences that they had shared with the new unions (e.g., class backgrounds, organizational cultures, 133 On the role of teachers in the PT s early organization-building, see interview with Hamilton Pereira in de Moraes and Fortes (eds., 2008: 265). 134 See Borges (1998): 96, cited in Secco (2011: 43). 135 Priest, author, ALN militante, early PT supporter, and CEB enthusiast. 112

121 commitments to equality, bottom-up change). 136 Yet the PT s socialism added appeal for many on the Catholic left, who, influenced by Liberation Theology, believed that Marxism and Christianity complemented one another. Betto famously claimed in 1986 that a real Christian is a communist, and a real communist a Christian (Secco 2011: 45). Members of CEBs and the Catholic left who participated in transition-era politics overwhelmingly joined or at least supported the PT. Frei Betto described the early PT as a vehicle for social and political transformation and the CEBs as the gasoline station. CEB members drew political inspiration from their CEB experiences and set out to find appropriate vehicles, from the PT and new unions to the MST and many other civil society actors. From the beginning, most CEBs operated in Brazil s urban periphery and countryside. At the time of the PT s formation, for example, thousands of CEBs operated in peripheral São Paulo alone. Because the Catholic left possessed a large grassroots base, CEBs, like the new unions, fed into the PT en masse. Many CEB members created PT núcleos de base. Some CEBs simply became PT núcleos. Tens, perhaps hundreds, of thousands of CEB members became PT members, activists, and leaders. Lincoln Secco recalled that most PT militantes in Greater São Paulo belonged to CEBs. Notable early leaders with CEB backgrounds included Olívio Dutra, Luiza Erundina, Marina Silva, and many other less well-known or unknown petistas. 137 Like the new unions, the Catholic Church provided important organizational resources, especially meeting places for PT nuclei and local committees. Both Lincoln Secco and a Catholic Church scholar, in interviews, stressed that, given the resource constraints of the PT and its base, local Catholic Church organs though they did not contribute financially to the 136 On CEB/PT affinity, see also Keck (1992: 97). 137 Interview with Frei Betto. 113

122 PT, even informally 138 provided meeting places for PT nuclei 139 and even offered courses in political consciousness-raising (conscientização política). 140 Because the Catholic Church, unlike the new unions, boasted a local infrastructure extending to the far reaches and forgotten corners of the national territory, CEBs significantly broadened the PT s early territorial reach, especially in rural regions. The Catholic Church s pastoral work in the Brazilian countryside was fundamental for the PT s achieving a truly national character (de Moraes and Fortes, eds., 2008: 262). According to Marxist PT founder Hamilton Pereira, the PT, through its linkages to the Catholic Church, [became] a party that [brought] along the various expressions of Brazil. Because the Church was] a national institution. 141 Lula himself has observed on multiple occasions that the Catholic Church played a more important role than unions in the PT s national implantation. 142 The PT and the Marxist left Brazil s Marxist left, from the beginning, had a complicated relationship with the PT. On the PT s formation, the Marxist left included a pragmatic, gradualist, and institutionalist Marxist establishment, comprising the country s two oldest and largest communist parties, the pro-soviet Brazilian Communist Party (PCB, est. 1922) and Maoist Communist Party of Brazil (PCdoB, est. 1962), as well as the MR-8, a PCB splinter and Brazil s third largest Marxist organization (Keck 1992: 96). 143 With the initiation of the abertura and distensão in 138 Interview with Secco. 139 Interviews with Secco and Almeida. 140 Interview with Rosemari Almeida. 141 Interview with Hamilton Pereira in de Moraes and Fortes (eds., 2008: 262). 142 Interview with Frei Betto. 143 The MR-8, a PCB splinter party, originated in 1964 under the name, the Dissidência Comunista de Guanabara, before taking the name, MR-8, in 1967 to commemmorate Che Guevara s death. The MR stands for Movimento Revolucionário, the 8 for the day Guevara died (October 8, 1967). The MR- 8 gained worldwide visibility after kidnapping the United States ambassador to Brazil, Charles Burke Elbrick. 114

123 the mid-1970s, the PCB had moderated, choosing to support the MDB, advocate a peaceful democratic transition, and seek positions of power within Brazil s slowly changing political regime (Secco 2011: 71). The PCdoB and MR-8 had adopted similar tactical positions as the military regime became more soft-line from the mid- to late 1970s. Given that the PMDB belonged to the center-left and, on creation, possessed national clout, all Marxist parties that chose to participate in institutional politics had an electoral incentive to join/support the PMDB rather than the PT. 144 Following on their previous support for the MDB, the PCB, PCdoB, and MR-8 joined the PMDB. 145 Not only did the Marxist establishment resist the PT, throughout the 1980s this establishment and above all the PCB vied with the new union movement and the PT for union loyalties. During the initial autonomous labor ferment of the late 1970s and early 1980s, new unionists had come to view the established Marxist labor leadership as insufficiently radical, willing to denounce peleguismo but not to frontally attack it (Secco 2011: 69). Secco (20110 summarizes that while in 1978, it was difficult to radically separate PCB unionists and future PT unionists, by 1980, differences between the new unions and the formal, Marxist union establishment had sharpened (Secco 2011: 67). The division between Marxists and the PT/new unionism culminated in 1983, when PT and new unionist leaders, in an effort to consolidate and strengthen autonomous labor after the PT s electoral setback of 1982, formally separated from the Unidade Sindical faction of Brazil s umbrella labor confederation CONCLAT considered peleguista by the autênticos and founded the CUT. Unidade Sindical had included new unionists and most Marxist labor leaders. The Marxists did not follow the new unions into CUT, opting to remain in 144 This fact distinguishes the PT from Mexico s PRD. In Mexico, the PMS had a strong electoral incentive to merge and did merge into the PRD because of Cuauhtémoc Cárdenas s national electoral clout. 145 The PCB s decision to ally with the PMDB has led some commentators to charge the PCB with failing to connect with the new spirit of the epoch (Rodríguez 2010: 146; see also Vianna 2001). 115

124 CONCLAT. Lincoln Secco summarizes that the PCB, until the mid-1980s, boasted a significant presence (capilaridade importante) in unions and remained a strong opponent of the PT within the union movement. 146 As a consequence of this competition, animosity and distrust between the PT and the non-petista Marxist left often ran high during the PT s early development. Among Marxists, for example, a rumor circulated widely that General Golbery do Couto e Silva, author of Brazil s National Security Doctrine, had supported the PT s development in order to divide the left and weaken the PCB within the labor movement (Secco 2011: 72). In a 1987 debate, Lula attested to the strength of Marxist parties union presence during the 1980s, asserting that the big delay in forming the PT were the Marxists precisely. 147 An internal PT booklet from entitled Os direitos dos trabalhadores e a Constituinte reflects ongoing tension between the PT and Marxist establishment, criticizing the communists for letting the PMDB co-opt them, and growing comfortable (acomodados) with third-rate government positions. 148 The above notwithstanding, elements of the Marxist left played a critical role in the PT s organizational development. Taken as a whole, the Marxist left varied considerably on whether to participate in institutional politics and, more centrally, 149 on which party, if any, to support. Although the Marxist establishment, as noted above, jumped on the PMDB bandwagon, many PCB, PCdoB, and MR-8 leaders and quadros defected and joined the PT 146 Talk by Lincoln Secco at PT Diretório Zonal in Pinheiros, São Paulo city. ( 147 From Socialismo em debate, : 259, a 1988 multi-author volume published my São Paulo s Instituto Cajamar. 148 From CSBH archive. 149 Only a small minority of Marxist groups opted not to participate in elections. 116

125 during the late 1970s and early 1980s (e.g., José Genoino, Paulo Fonteles, Manoel de Conceição, Jamie Santos). The PT also attracted numerous, more radical Marxist organizations and networks. The Trotskyist Fração Operária (Workers Faction), for example, dissolved into the PT wholesale. The PT, in order to maximize its appeal to on-the-fence radical Marxists, tacitly permitted dupla militância, whereby petistas would enjoy license to remain active in and primarily committed to other parties (about which more later). Consequently, a group of small, radical Marxist parties (e.g., Democracia Socialista, PCBR, APML, and the MEP) joined the PT while remaining independent and explicitly treating the PT as a front for their revolutionary long-term goals (Keck 1992: 96). 150 Other Marxist groups did not join the PT but actively supported it, given the PT s external origins and ties to autonomous labor. Unlike the new unionists and Catholic progressives, the Marxist founders of the PT did not provide a large social base. In 1991, only ten percent of petistas belonged, or had belonged, to Marxist organizations of the extreme left (Secco 2011: 48). These organizations were small, vanguardist networks, many having engaged in clandestine activity since the 1960s (Ribeiro 2010: 186). Yet they punched above their weight by contributing abilities and characteristics otherwise in short supply. Due to Marxists internal discipline, education and debating skills, mobilizing experience, and capacity for activist training, PT radicals (relative to moderates ) exercised disproportionate influence within the early PT organization, whether in shaping the PT s early ideology and program, 151 organizing and leading the first base-level 150 The PT s Marxist elements considered the PT little more than an electoral front, or a space to be occupied tactically, with a long-term view toward building a revolutionary party (de Moraes and Fortes (eds., 2008): 32). 151 Many petistas focus on the education and rhetorical skill of the early PT s Marxist leaders and quadros. These characteristics enabled the Marxist left to influence the PT s ideological development, and also en- 117

126 PT núcleos and offices, 152 training PT activists, 153 advising PT elites, 154 or assuming major national leadership positions. 155 Still, [t]he importance of Marxists in the PT was always greater internally (na disputa interna) than externally (Secco 2010: 49). Although Marxists played a critical role in early organization-building, they possessed a limited ability to sway individuals outside the party abled many individual PT Marxists to win positions of influence on the equipes of non-marxist PT elites. Keck (1992) writes that early Marxist petistas maintained an impressive level of militancy and...seemed exceptionally articulate (95). One early petista and campaign consultant summarized that the Marxists, unlike the new unionists and left Catholics, possessed a detailed program (interview with Sergio Alli). A Marxist ex-petista made the general observation that Marxists, given their education and skill in argumentation, often dictated internal negotiations with the new unionists, both nationally and subnationally (interview no. 3). Given the general weakness, chaos, and deficiency of the PT s internal press (Secco 2011: 108), the party s Marxist networks, with newspapers like Em Tempo and O Trabalho, dominated internal public communications (Secco 2011: 110). 152 Several members and analysts have emphasized the Marxists role in mobilizing local PT networks and creating local and state party organs. Ribeiro (2010) stresses the Marxist left s high capacity for mobilization and activist regimentation (251). Two individuals, one a PT founder and leader, the other a younger Marxist ex-petista, summarized that Marxists at the time of the PT s formation possessed invaluable experience in political mobilization, which enabled them to play a disproportionate role in member recruitment and training at the local level and in the creation of local and state Diretórios (interview with Azevedo and interview no. 3). Marxist parties, networks, and cliques played an especially important party-building role in urban areas, where most lived, had received an education (often a college education), and could lead, mobilize, recruit, educate, and train individuals in a variety of local movements and organizations, from student and neighborhood associations to the new unions. 153 Marxists brought with them an emphasis on activist and leader conscientização (consciousness-raising) and training. Djalma Bom recalled that the Marxist tendências conscientizou many petistas, including himself (interview with Bom). One Marxist ex-petista observed that the Marxists influenced the popular leaders (e.g., new unionists) as much as the popular leaders influenced the Marxists: O marxismo se concretizou e o concreto se marxizou (interview no. 3). In 1986, the PT, at the insistence of Marxist petistas, created a centro de formação política, the Instituto Cajamar (de Moraes and Fortes, eds., 2008: 282), and mandated, with limited success, that ten percent of party funds go to political training. Secco (2011) writes that in practice, the PT never fully adhered to this internal statute. 154 Secco (2011) observes that many Marxists became formuladores for new unionist PT elites, although they often had to tone down their ideological views in order to hold these positions (48). 155 Due to their political experience and high education levels, Marxists also assumed a disproportionate number of major national leadership positions. Half of the PT s first National Executive Committee (ten of twenty) were new unionists (Secco 2011: 64), and most of the remainder were Marxists. (Most of the new unionists in the CEN belonged to middle-sector unions, not industrial unions, further evidencing the importance of education for attaining the highest positions within the PT (Secco 2011: 64).) While a new unionist, Jacó Bittar, served as the PT s first national secretary of organization, most members of Bittar s office belonged to the Marxist left (e.g., José Dirceu, José Ibrahim, Jorge Baptista) (interview with Azevedo). 118

127 fold. For this reason, Secco (2011) writes that the CEBs and new unionism were the two most important social vectors of the PT s formation (49; emphasis not in the original). In places where new unions played the dominant role in early PT organization-building, both the Marxist and Catholic left typically played important secondary roles (e.g., the Pastorais Operárias in the ABC paulista; the Trotskyist Democracia Socialista (DS) in Porto Alegre). Moreover, and crucially, where new unions did not have a strong presence, the Catholic left and Marxist left played the leading roles, in some cases sharing the responsibilities of organization-building. Melo (1994) observes that in developed urban localities dominated by traditional unions (as distinct from new unions), the Marxist left and affiliated movements/organizations typically spearheaded the party-building process, while in rural localities, the Catholic grassroots left often led the way. In Rio de Janeiro, Belo Horizonte, Fortaleza, and Londrina, where most unions did not affiliate with the new unionism, Marxist intellectuals and activists, in conjunction with other groups, took responsibility for recruiting members and setting up local offices (Keck 1992; Melo 1994). In more rural states such as Acre (Keck 1992), 156 Pará, 157 Paraíba, 158 Piauí, 159 and Goiás (Ribeiro 2010: 251), CEBs associated with the Pastorais da Terra played the leading role. In Vitória, the capital of Espírito Santo, CEBs played the leading role. In Pará, Goiás, Paraíba, and Rio Grande do Norte, the Marxist left made important contributions, especially in supplying top PT officials (e.g., the PCdoB and MR-8 in Pará) As of the early 1990s, Acre was the only Brazilian state in which the PT had achieved full organizational penetration at the municipal level (Keck 1992: 101). 157 Interviews with Avelino Ganzer and Manoel de Conceição in de Moraes and Fortes (eds. 2010). 158 Secco (2011: 43-61). 159 Secco (2011: 43-61). 160 You know that the PCdoB tried to build the PT in some states. That s what happened here [in Pará]. So much that the president of the provisional committee was Paulo Fonteles of the PCdoB, the secretary general was a guy from MR-8, Jamie Santos (interview with Paulo Rocha in de Moraes and Fortes, eds., 2008: 148). On Marxists in Rio Grande do Norte, see Secco (2011: 43-61). 119

128 * * * The PT benefited from ties to other civil society organizations and movements as well. From the mid- to late 1980s in particular, the PT s societal linkages broadened from its original bases, as the party managed to absorb a large part of [the] energy 161 associated with the Diretas Já! movement (1983-4) 162 and the MST (est. 1984) (about which more later). 163 Landless workers, for example, played a key role in building the PT in Mato Grosso and Mato Grosso do Sul. In addition, the PT drew from neighborhood associations in Rio de Janeiro, 164 the health movement in Sorocaba, São Paulo, public transportation and environmental activists in Cubatão, São Paulo, textile workers in Paraíba, favela associations and public transportation users in Santo André, São Paulo, and neighborhood associations and post-materialist movements in several cities (Secco 2011: 43-61). These additional civil society ties notwithstanding, the PT, for early organization-building, primarily depended on the new unions, the progressive Catholic Church, and the Marxist left. Individuals and networks from these three groups performed the large bulk of early organizational work. Early territorial expansion tended to occur in the areas, predominantly but not exclusively urban, where the new unions, Catholics, and/or Marxists had a strong presence. Elsewhere, the early PT, in the typical case, did not become heavily implanted. Building a mass organization A large membership and infrastructure do not, in themselves, make party organizations strong. What made the PT strong was the size but also, crucially, the commitment of 161 Phrase from talk by Lincoln Secco at PT Diretório Zonal in Pinheiros, São Paulo city. 162 Secco 2011: Interview with Hamilton Pereira in de Moraes and Fortes (eds., 2008): Benedita da Silva, future PT governor of Rio, became involved in the PT through her leadership in Rio favela associations. 120

129 its base. The PT s adverse context of formation not only created an incentive for party leaders to invest in territorial organization, it also selected for early party-builders, or early joiners (Greene 2007), with high levels of commitment. Consequently, the early PT built an organization with impressive territorial reach, composed of believers (Panebianco 1988: 26-30). The size of the early PT s infrastructure and base The early PT established a mass activist base and extensive local infrastructure in multiple regions of the country. Nuclei, Diretórios, and dense activist networks sprouted up in the industrial and urban municipalities of the Southeast (e.g., São Paulo, Minas Gerais) and South (e.g., Rio Grande do Sul, Espírito Santo) and in select rural states (e.g., Acre, Pará) due to the concentrated presence of civil society feeder organizations. PT membership shot up during the period. In May 1980, the PT counted roughly 30,000 members (Rodríguez n.d.: 231). By June of the following year, the number had multiplied seven fold, to approximately 200,000, or 0.36 percent of the national votingage population. By July, the PT had organized state Diretórios in twenty-one states and claimed more than 20,000 members in four states: São Paulo, Minas Gerais, Rio de Janeiro, and Rio Grande do Sul (Secco 2011: 50). On Secco s (2011) summary, at the time of the 1982 elections, the PT was in the main states and already boasted a militância even in the Northern region, more distant from Brazil s major political and economic centers (50). As the PT set out to fulfill the legal requirements in , the national leadership, alongside the formal party organization, actively encouraged the development of informal participatory nuclei something not mandated by law. From 1980 to 1982, the period of greatest base-level PT ferment, thousands of petista activists created núcleos de base: informal baselevel networks, composed of at least 21 petistas, which met in houses, garages, churches, etc. 121

130 to discuss, debate, and plan party-related matters. The nuclei did not receive financial aid from the PT and often sold party paraphernalia to finance their activities. In many cases, PT nuclei succeeded base-level structures linked to PT feeder organizations, especially CEBs. 165 The PT s conception of the nucleus...had an affinity with the organizational form of the base communities of the Catholic Church, especially due to the profoundly anticentralist bias underlying both organizational forms (Keck 1992: 104). Nuclei mimicked the CEBs, according to Secco (2011: 78). Keck (1992) finds an association between rates of nucleus formation and CEB density (104). Núcleo membership peaked around five percent in 1982 but fell precipitously over the course of the 1980s (Secco 2011: 84, 263). Many núcleos dissolved into campaign committees and local Diretórios. 166 While in 1982 the PT boasted roughly 1000 núcleos with nearly 30,000 individuals participating (Secco 2011: 80), the number of núcleos had fallen below 700 by the mid-1980s (Ribeiro 2010: 263) and to 200 by the mid-1990s (Ribeiro 2010: 87). After , the PT s membership and formal infrastructure expanded in a relatively steady manner. At 0.36 percent in 1981, the PT s membership as a proportion of the voting-age population rose to 0.44 percent in 1984, to 0.60 percent in 1988, and to 0.74 percent in 1995 (Ribeiro 2010: 244). With regard to party infrastructure, by 1985 the PT had organized municipal Diretórios in 1100 of 4022 Brazilian municipalities (Secco 2011: 117). During the second half of the 1980s, the PT, largely due to the CUT s expansion into rural and new urban areas, made considerable organizational advances outside São Paulo state. By 1989, the PT had established a municipal branch in nearly half (44 percent) of Brazil s municipalities. In absolute terms, in 1989, the PT boasted 625,000 members (Keck 1992: 165 Interview with Frei Betto. 166 Confusion over specific functions of nuclei within party, coupled with precarious intraparty communications, tended to erode nucleus formation over time. Many nuclei were absorbed into electoral committees in 1982 and never reconstituted (Keck 1992: 104). 122

131 110). By the mid-1990s, the number approached one million. Adversity, selection effects, and the commitment of the PT s early joiners Many petistas characterize the party s organizational push during the 1980s as an undertaking of Himalayan difficulty, achieved through force. Primarily, the difficulty stemmed from the need to recruit hundreds of thousands of new members. Given the PT s low media access, early organizers had to engage in person-to-person (corpo a corpo) recruiting, and given the party s lack of resources, they could not offer selective benefits. Instead, they had to inspire individuals to devote their time and labor to the nascent PT. Further, in recruiting petistas, party-builders had to compete both with the MDB/PMDB, which urged Brazilians not to waste their vote on the PT, 167 and with the PCB, which boasted strong ties to sectors of organized labor and considered itself the authentic party of the Brazilian working class. 168 When recalling the initial years of development, petistas often stress the difficulty of satisfying the legal organizational requirements, 169 and the negative electoral consequences of the voto vinculado. 170 Keck (1992) writes that the PT gained legal status in 1980 against seemingly impossible obstacles (Keck 1992: 93), referring to the legal organizational requirements for registry. Paulo Rocha, PT founder from Belém, Pará, recalled that in Pará, we legalized the PT almost by force (meio na marra), with the creation [of offices] in eighteen municipalities. 171 The voto vinculado law, in effect for the 1982 elections, slowed the PT s electoral progress by forcing the party to run viable legislative candidates in executive 167 The PMDB launched a voto útil campaign against the PT in São Paulo in Interview with Djalma Bom, published in de Moraes and Fortes (eds., 2008: 90). 169 E.g., Interview with Djalma Bom in de Moraes and Fortes (eds., 2008: 89-90). 170 Interview with Raul Pont in de Moraes and Fortes (eds., 2008: 226). 171 de Moraes and Fortes, eds., 2008:

132 elections that they stood no chance of winning. 172 Rodríguez (n.d.) summarizes that the PT s early organizational development involved an extraordinary effort to incorporate distinct groups and sectors across an extensive territory (Rodríguez n.d.: 225). Two founding petistas offered especially vivid accounts of early difficulties. First, in an unprompted statement from a published interview, Djalma Bom observed that although the conjuntura of the late 1970s facilitated and contributed to the creation of the PT, the hardships were immense [emphasis Bom s]. You cannot imagine how hard it was to create the PT. 173 Later in the same interview, Bom reiterated the point: You cannot imagine the difficulties of forming the PT. [People] thought we wouldn t meet the legal requirements. But the thing went forward, with all the difficulties we had... The legislation itself militated against (cerceava) the creation of political parties. We had to redouble forces (nós tivemos que nos desdobrar) in order to satisfy [them]. 174 In the state of São Paulo alone, the PT, by late 1980, had to establish more than one-hundred municipal offices in order to achieve registry: We would travel into the countryside, without anything, with an old Volkswagen Beetle, and we did not have money. During that period, at midnight, the tolls opened at midnight allowing us to pass for free, so we would wait... Everything was very difficult. 175 One of the initial MDB defectors, Irma Passoni, described her 1982 campaign for federal deputy in similar terms: I didn t even spend much money. The little debt I accrued I paid off in the following term. The political campaign was assumed by all the militantes. There were people who made notes on papel de pão and gave them to people... It was really complicated, involving a lot of sacrifice and suffering as well. The difficulties increased when I was the PT s secretary of Organization and I went to create the party s provisory committees. If it was hard here in the capital, it was much worse in the interior. We were called communists, revolutionaries, I don t know what. How were we going to organize the PT in the 120-something municipalities necessary [for legalizing the party]? It was an insane task. The initial construction of the PT in the interior was done by workers who had left the ABC and by members of the community. I was 172 Anticipating this, an internal tactical paper from Belo Horizonte, Minas Gerais in Christmas 1980 stresses: [W]e seek to elect lots of municipal councilors, some mayors, and to avoid the defeat of deputy candidates without a chance of winning, who would stand a reasonable chance of winning for the PT if they ran for municipal councilor or mayor ( Eleições em debate: Uma tática eleitoral para o PT, from CSBH archive). 173 de Moraes and Fortes, eds., 2008: de Moraes and Fortes (eds., 2008): de Moraes and Fortes (eds., 2008: 90). 124

133 secretary of Organization without any estrutura, without a cent. We slept in the homes of community participants, we ate in their homes, and they often had little to eat. That s how we built the PT. 176 For the PT, the difficulty of early party-building also stemmed from the repressive atmosphere of the late 1970s and 1980s. Unlike the PMDB, the PT and its feeder groups (both before and after the PT s formation) originated in a hostile context (Rodríguez n.d.: 225). During the 1970s and 1980s, the Marxist left, autonomous urban unionists, and rural workers at different periods, in different regions, and to varying degrees suffered repression at the hands of government forces or local oligarchies. Although the general public often did not know about or give special salience to the repression of workers, the PT did. Campaign materials from the 1980s demonstrate that PT leaders frequently invoked the repression of workers in order to galvanize and grow the base. Marxists suffered much of the torture and murder under hard-line military rule. Autonomous urban labor, in contrast, suffered very little extreme repression. Military and police killed a relatively small number of urban new unionists from the mid-1970s to the mid-1980s (e.g., Manoel Fiel Filho in January 1976; two strikers in Leme, São Paulo, soon after the transition to civilian rule). More often, government forces intervened in unions to remove the leadership or broke up strikes, hurling political epithets at workers and their collaborators and arresting union leaders. In 1979, the Figueiredo government arrested dozens of top new union leaders, including Lula, and detained them for weeks in April/May In response, Lula and fellow new union leaders carried out a hunger strike, leading to their eventual release. During the PT s formative phase, no feeder group suffered greater repression than autonomous rural workers and the actors who helped organize them (above all, the Catholic 176 Passoni continues: That s also what happened in the national organization of the PT. The available estrutura was that of the mandatos of the deputies... We depended on militantes at the regional level: in Piauí, Ceará, Rondônia, Acre... And that s how we built the PT in all of Brazil, with the efforts of each individual, in each place, with each person s resources. It was born from the forces that each person possessed... (interview with Irma Passoni in de Moraes and Fortes, eds., 2008: 318-9, 321). 125

134 left). Like their urban counterparts, autonomous rural workers frequently faced interventions and verbal abuse from military, police, and local elites. Yet for rural workers, forceful interventions only scratched the surface. In rural Brazil, local oligarchies enjoyed near total impunity due to the weakness of civil society, media, and the state. Between 1975 and 1989, these oligarchies, sometimes aided by military and police, murdered 1377 rural workers in agrarian conflicts (Petit 1996: 142). 177 Murders of rural workers spiked during the Constituent Assembly, when the push for land reform intensified. PT leader Hamilton Pereira recalls: It was the period in which we were mobilizing to interfere in the Constituent Assembly, , that the degree of violence in the land struggle climaxed (chega ao paroxismo). Every third day a union leader, a nun, a priest, someone involved in the land struggle was killed. 178 In the rural states where the early PT established a strong presence (e.g., Pará, Acre), the PT necessarily reflected these struggles between rural workers and local oligarchies (Secco 2011: 46; also see Petit 1996 on the PT s development in Pará). The adverse conditions of the PT s formation (scarce resources, low media access, repression, disadvantageous electoral laws) created powerful selection pressures. The party s lack of access to resources and media and steep organizational barriers to entry made early partybuilding non-remunerative and highly labor-intensive. Repression raised the cost of participation for certain feeder groups, especially rural workers in the Northern region and certain Marxist networks. These difficulties, along with the party s dim short-term electoral prospects at the national level, weeded out political opportunists and selected for true 177 In addition to the murders, local oligarchies and allied military and police frequently intervened in rural worker strikes and demonstrations, displaying force and verbally assaulting both union organizers and the poor, often illiterate rank-and-file. A rural CUT leader, PT founder, and senatorial candidate recalled a representative episode in which military officers halted a mass demonstration of autonomous rural workers in Rurópolis, Pará in May 1980: We were surrounded and monitored by military personnel on all sides... It seemed like the end of the world because there was enormous persecution. They invented things that I didn t understand... When they said that we were communist, I didn t know what communism was. I didn t understand anything (de Moraes and Fortes, eds., 2008: 170). 178 Interview with Hamilton Pereira in de Moraes and Fortes (eds., 2008: 266). 126

135 believers, or the ideologically committed. Given the low short-term benefits and often high short-term costs of participation, individuals without strong ideological convictions, by and large, did not join the PT, even if they leaned left politically. PT members and observers have noted as much. Bruhn (2008) treats the PT s early development as comparable to that of Mexico s PRD. Citing Greene (2007), she writes that both were antisystem parties, founded in the course of a struggle against an authoritarian regime and with the explicit purpose of confronting that regime... Such parties are of interest only to people with extreme views who are driven by ideology more than the desire for power... They are natural-born protesters (165; emphasis not in the original). Djalma Bom recalled in an interview that the PT militância used to be warlike (aguerrida)... Things were much harder, but we did things with much more happiness and solidarity (integração) (de Moraes and Fortes, eds., 2008: 95). In a 2008 interview, Olívio Dutra, one of the PT s central founders, stressed, [a] party that enters the government runs the risk of...rapidly attracting opportunists of all stripes who want benefits. 179 The PT, he observed, did not arise from public office, either legislative or executive, but from the struggle of the Brazilian people (de Moraes and Fortes, eds., 2008: 124). A PT founder and campaign consultant similarly observed that while most Brazilian parties depended on clientelism, the PT, with no executive power, relied on a social base driven by conviction, not material incentives. 180 But what conviction drove early petistas? In order to fuel local party-building efforts, party leaders must provide some type of incentive to activists, and if leaders cannot distribute selective material incentives (jobs, money), they must offer collective incentives: an identity and/or set of goals shared widely by party members, which transcend individuals desire for wealth, power, and status (Panebianco 1988: 24). 179 de Moraes and Fortes (eds., 2008):. 180 Interview with Alli. 127

136 Petistas did not rally around a single positive ideology (about which more below). Members and feeder networks ran the gamut of left ideologies, from orthodox Marxism to pro-capitalist social democracy. Equally, petistas did not coordinate around the goal of ending the military dictatorship. If the desire to end military rule had fundamentally animated petistas, they would have joined the PMDB, which also opposed the military regime, advocated a swift end to military rule, and, crucially, possessed far greater electoral clout than the PT. If not united by a specific, positive ideological vision, or by simple opposition to the military dictatorship, what positive goal and/or identity did early petistas coordinate around? Why were such a diverse range of individuals willing to devote time and labor to an electorally marginal party-building project? From the PT s inception, PT leaders branded the party, and members viewed the party, as a novelty in Brazilian history: the first authentically popular expression in a society perpetually dominated by elites. 181 The PT s popular origins distinguished it from all other Brazilian parties, including the opposition PMDB. During the PT s early years, campaign literature explicitly lumped the PMDB in with the rest of the bourgeois political establishment. Petistas characterized the PMDB as a domesticated and artificial opposition, cut from the same elite cloth as all the other parties. A 1982 flier for a group of PT candidates from Rio de Janeiro, for example, described both the PDS and the PMDB as the parties of the landowners and factory owners. 182 In contrast, the PT, to quote one campaign booklet, was oposição pra valer ( an opposition that counts ). In an illustrative quotation from a 2008 interview, PT leader from Minas Gerais, Luiz Dulci, explained why the PT, in its earliest days, did not dissolve into, or at least ally with, the progressive Tendência Popular of the MDB: 181 Interview with Alli and interview no From the CSBH archive. 128

137 Look, it wasn t due to differences in political platform. There were differences in that respect, but they were surmountable (contornaveis). In my opinion, the root problem, even if no one recognized it openly, was the forma de fazer politica. When asked, In what sense?, he replied that the parlamentares of the MDB did in fact have deep democratic convictions and sincere social commitments, but theirs was a traditional vision of the party...with an exclusively institutional conception of politics. They had causes in common with Lula and with us [the PT], but they did not give the same weight as we did to social movements, nor to the whole of civil society; they didn t think it was possible to create a party from the bottom up, truly participatory, in which the popular classes acted directly and above all became politically educated in order to lead the state one day. We wanted more than to restore representative democracy...we wanted to give [representative democracy] a new quality, making each citizen, in their daily life, an active political subject. 183 Early campaign materials focus on the popular character of the PT, not the characteristics of individual PT candidates, who often shared platforms verbatim and ran unitarily. Countless homemade and party-produced pamphlets and fliers emphasize that the PT s distinctiveness lay in its popular origins. These texts claimed that political change could only come de baixo para cima (from the bottom up), and that the PT alone involved the masses in the political process. A 1982 campaign flier for Roberto Martins, PT candidate from Bahia, stated that [a]t this electoral juncture, the PT distinguishes itself from the other opposition parties, calling all the people to organization and struggle. 184 Early PT slogans included, A worker votes for a worker, and the Portuguese rhyme, Vote number three [the PT s ballot number], the rest is bourgeois. 185 The PT s authentically popular origins gave it a mística that resonated across the left spectrum, including with Marxists. During the initial years especially, the PT s unique sociological profile attracted a wide range of left groups, including many on the Marxist left. 186 In the words of one Marxist ex-petista, the PT s founding myth of popular authenticity permitted incredible diversity. 187 Marxist PT leader Raul Pont recalled in a 2008 interview 183 Interview with Luiz Dulci in de Moraes and Fortes (eds., 2008: 204). 184 From the CSBH archive. 185 From CSBH archive. 186 Personal communication with Ribeiro. 187 Interview no

138 that although Marxists did not agree with many of the PT s positions, the PT was the party that Marxists had been waiting for, dreaming about : [In the 1960s], we wanted [a party with a real social base] but there were no workers, no laborers, not even masses. [There was a] ferrada repression, a smaller working class, unions with pelêgos... So we said: Look, the PT is the party we dreamed of, the party we always fought for, a party of the workers, so we don t need to remain outside. We don t need to fight for the complete line today, tomorrow. It s a process. What s important is for the party to be of the workers, class-based. Our role here is to fight to make it internally democratic, anti-capitalist, protective (defensor) of democratic socialism... Although the PT would spend the 1980s almost entirely out of government, the PT s governing ethos, established later in the party s history, would reflect a continued fundamental emphasis on popular participation and empowerment. When the PT began to win important positions in the late 1980s and 1990s, party elites trumpeted the modo petista de governar, which centered on popular participation, over any particular ideological or programmatic goal (Samuels 2004). The party s most innovative and influential experiment in municipal governance (Porto Alegre, ) achieved world renown due to participatory budgeting, not to specific social policies such as education or health reform. In sum, from the beginning, the PT s popular identity, and the goal of putting the reins of government in workers hands, animated petistas and, crucially, extended their time horizons. Petistas sought to transform Brazil s historically elite-dominated political system and knew that success, given the PT s utter marginality in the 1980s, would require dogged persistence. Accordingly, party leaders explicitly downplayed the significance of particular, short-term electoral victories, keeping their gaze fixed on radical political transformation. Describing his 1982 campaign for governor of Rio Grande do Sul, for example, Olívio Dutra stated: We had a disposição incrível. A very spontaneous thing... We clearly understood that [the public] machine and the state were responsible for all the inequalities, the injustices, and the situation that we had to change. There was a lot of heart, excitement, a very rooted conviction, due to the social struggles, that the people had to play the leading role in history and that 130

139 the PT had to carry this proposal forward, independently of whether we won the elections (emphasis not in the original). 188 Keck (1992) observes that after the disastrous elections of 1982 (about which more below), petistas rededicated themselves to union and social movement activity, as if the election campaign had represented a deviation from the normal goals of party activity (153; emphasis Keck s). A leading early petista from Minas Gerais recalled that [w]e wanted a mass party but organic, activist, not just a party of public opinion. A party that didn t just operate during election season. 189 In a 1986 internal PT interview, party leader Gumercindo Milhorem Neto echoed an oft-voiced sentiment: Winning an election doesn t mean winning power. Power in São Paulo, in the cities and in the states in general, belongs to the large property owners: the owners of car factories, of bus businesses, of transportation businesses, the big ranchers, the big bankers; these are the powerful. 190 The PT s 1989 presidential campaign reflected the party s continued non-electoralist character, as party leaders knowingly took electorally suboptimal stands on potentially decisive issues (Hunter 2010: 112). In particular, Lula and the national elite remained explicitly socialist, chose not to pursue an alliance with the PMDB, and refused to condemn the increasingly aggressive MST, recognizing that such decisions would provide fodder for the conservative opposition (Hunter 2010: 112). The above notwithstanding, the non-downsian character of the early PT should not be overstated. The internal life of the PT, almost from the beginning, did revolve around elections. As noted above, the party s main grassroots structures for dialogue and debate outside of elections, the núcleos de base, only ever involved a small minority of party members and weakened considerably very early in the party s history. Secco (2011) argues that [t]he 188 Interview with Olívio Dutra in de Moraes and Fortes (eds., 2008: 124). 189 Interview with Luiz Dulci in de Moraes and Fortes (eds., 2008: 204). 190 Published interview entitled, A Dívida É Fabricada (1986). From CSBH archive. 131

140 ephemeral nature of the nuclei had an obvious cause. Most activists got excited about (se empolgava com) electoral campaigns (Secco 2011: 85). After a couple of years, he states, elected officials and [t]he leaders of tendências and union organizations made careers, and the bases went home (Secco 2011: 77). Keck (1992) summarizes that, by the late 1980s, the PT s desire to maintain active grassroots party organizations outside of election periods ha[d] gone largely unfulfilled (121). Organizational expansion and electoral progress Most of the PT s electoral progress during the 1980s followed from organizational expansion, at the state and federal as well as municipal levels. The PT could only run municipal candidates where it had a diretório, and for federal, state, and local candidates alike, the presence of a local party organization was important for mounting a campaign (Keck 1992: 152). Activist networks kicked into gear for elections at all levels. Municipal and zonal Diretórios served as local campaign headquarters (bases de campanha). 191 Given the PT s lack of access to media and resources, the PT s votes, throughout the formative phase, tended to come either from PT militantes or from the persuadable voters they managed to reach in person. Consequently, in electoral terms, the early PT performed best in its organizational bastions (e.g., the ABC paulista). Keck (1992) finds that in the São Paulo state election in 1982, the PT received less than two percent of the vote in 97 percent of municipalities without a PT Diretório, and over five percent of the vote in 57 percent of municipalities with one (152). Meneguello (1989) finds, more generally, that throughout the 1980s, the PT s electoral results closely tracked pre-election organizational strength. In places with local PT offices and large PT memberships, the PT performed well, independently of other factors, even rally attendance. 191 Interviews with Venturi and de Filippi. 132

141 The PT s slow progress in large executive elections, far from reflecting organizational weakness, demonstrated that even an exceptionally strong grassroots organization could not compete with the media empires and patronage machines at the disposal of PT opponents. One party founder and intellectual recalled that for major executive elections, the Horário Gratuito alone mattered more than the militância. 192 Yet in the smaller elections, especially municipal elections and legislative elections at the state and federal levels, corpo a corpo campaigning played a key role often a fundamental, decisive role. 193 Consequently, the PT s performance in these lower-level elections tracked the party s steady organizational progress, both in traditional strongholds and in parts of the national territory where the PT developed new linkages with autonomous workers and other civil society actors. In its first decade in the Chamber, for example, the PT experienced an incremental growth in support (Keck 1992: 160), progressing from eight (1982) to sixteen (1986) to thirty-five (1990) elected deputies. Just as the PT s organizational strength contributed to electoral success, election campaigns and electoral success contributed to the PT s organizational strength (Keck 1992: ; see also Bartolini 1983). Save the PT s first major organizational push, when party-builders set out to satisfy the legal organizational requirements for registry, the periods of greatest organizational expansion occurred both in the lead-up to and wake of elections. In large measure, organizational expansion (particularly membership increases) during and after elections resulted from the PT s heightened visibility and following the 1985, 1986, and 1988 cycles increased electoral credibility. Equally, however, the PT intentionally campaigned to organize (Keck 1992: ). Party leaders consciously used elections, especially large executive elections (e.g., presidential, São 192 Interview with Azevedo. 193 Interviews with de Filippi and Azevedo. 133

142 Paulo gubernatorial), to enlarge the permanent party organization, capitalizing not only on heightened visibility and increased credibility, but also on campaign infrastructure. In the lead-up to the 1982 elections, for example, José Dirceu wrote that the electoral campaign [of 1982] will serve for us to expand the organized bases of the PT, doing recruitment drives and nucleus formation simultaneously. 194 Also in the lead-up to the election, the São Paulo state Diretório s Secretary of Press and Publicity (Divulgação) emphasized in a campaign strategy proposal that the grupos de apoio for the 1982 elections should become tomorrow s PT nuclei. 195 The trend continued for the remainder of the decade (and beyond). After the 1985 mayoral elections, the PT carried out recruitment drives (Keck 1992: 109, 156). A Manual de Grupo de Apoio for the 1986 congressional elections, written as a dialogue, asks whether support groups can become permanent after the campaign and answers yes, providing relevant instructions. 196 In a letter to PT militantes before the 1986 elections, Jorge Baptista emphasized that [t]he electoral campaign must serve the organization and consolidation of our party: the PT. 197 A flier supporting João Antônio for state deputy in 1986 read:...[w]e choose to run a candidate, envisioning the strengthening of the PT and the formação of new militantes, so that the party contributes to the advancement and politicization of popular movements and unions. 198 A flier for Luiza Erundina s 1988 São Paulo mayoral campaign reads: The candidacy of Luiza Erundina synthesizes the PT s objectives in the electoral progress: win the elections and, beyond that, impel party-building (impulsionar a construção 194 O PT e as eleições de From the Edgar Leuenroth archive. 195 A campanha eleitoral do PT: Divulgação e propaganda uma proposta de discussão. From the Edgar Leuenroth archive. 196 From CSBH archive. 197 From the Edgar Leuenroth archive. 198 From Edgar Leuenroth archive. 134

143 partidária), organize and conscientizar workers and the general population. 199 Organizational strength and the survival of crisis Not only did organizational strength enable the PT to make steady (albeit slow) electoral progress in an adverse context, the PT s grassroots capacity and ethos fortified the party amid early electoral crisis. The PT suffered its most crushing setback very early, in the 1982 municipal, state, and congressional elections. The party had entered the electoral season, its first, with great optimism due to bottom-up fervor, the great energy of struggle (energia grande de luta), 200 and the unexpectedly large rally audiences associated with the PT s 1982 campaigns (Keck 1992: 141-4). In particular, given Lula s national profile and strong performance in the 1982 gubernatorial debates, petistas approached the São Paulo gubernatorial election with high hopes. Most activists believed that Lula would win the governorship, arguably the second most important office in the country, and certainly the most important of the offices contested in Everyone who participated...in the electoral campaign, recalled 1982 federal deputy candidate Djalma Bom (São Paulo), expected [Lula to be elected governor]. And it wasn t for nothing because the largest rallies were done by the PT and Lula... That created the expectation, you know?. 201 Yet Lula placed a distant fourth in the São Paulo gubernatorial race. Olívio Dutra, probably the second most important leader in the party, placed last in the Rio Grande do Sul gubernatorial race. The PT did elect municipal councilors (vereadores) in local strongholds across the country, in São Paulo state, major urban centers in and out of the South and Southeast 199 From the CSBH archive. 200 Interview with Donato. 201 de Moraes and Fortes (eds., 2008):

144 (e.g., Bahia), and even remote, rural states such as Acre, Amazonas, Pará, and Rondônia (Secco 2011: 50). Yet in a country with more than 4000 municipalities, the PT won only two mayoralties, in the small cities of Diadêma, bordering São Bernardo do Campo, and in Santa Quitéria do Maranhão. 202 In the Chamber of Deputies, the PT elected a paltry eight legislators, three-and-a-half percent of the total. Keck (1992) calls the 1982 results a profound shock and disappointment to the PT (Keck 1992: 149), which the PT experienced as a severe defeat (152-3), and after which deep disappointment and a kind of collective depression set in (156). In the post-election analysis, party leaders concluded, among other things, that the 1982 campaigns had distanced the party from its civil society roots. The party initiated a return to the base and a renewed emphasis on social action (Keck 1992: 197), both in the organized labor sector and more broadly (Diretas Já!, the landless workers movement). Unbeknownst to PT leaders, this reinvestment in civil society would reap electoral dividends just a few years later. With no elected offices to occupy, top party leaders (e.g., Lula, Olívio Dutra, Jacó Bittar), rededicated themselves to the autonomous labor movement. As discussed earlier, in 1983, they founded the CUT, beginning a highly successful effort to broaden and strengthen the new unionism over the next few years. The PT s renewed focus on social action quickly came to involve other civil society actors and movements as well. In 1984, a set of regional, previously uncoordinated landless worker movements joined to create the national Landless Workers Movement (MST). Land reform had already become an important element of the PT platform and would only become more important as the 1980s progressed, as the movement strengthened and the push for land reform intensified. Although the MST eschewed partisan involvement, landless workers and rural activists predominantly supported the PT 202 One PT founder and intellectual attributed the PT s success in Diadêma, interestingly, to two factors: the strength of the PT militância, and the fact that Diadêma did not have a single TV or radio station (interview with Azevedo). 136

145 as of the mid-1980s. By the late 1980s, the PT [had been] affirmed as the land struggle s main partisan reference, especially given Luis Carlos Prestes exit from the PCB. 203 The PT s involvement in Diretas Já!, however, played a more central role in the party s post rebound. The period gave rise to Diretas Já!, the largest social movement in Brazilian history. The movement arose in early 1983 amid economic crisis. 204 A combination of economic slowdown and extreme inflation eroded the military s legitimacy, and anti-regime sentiment intensified. A wide range of actors, from political parties to movements and civil society organization, joined to press for direct presidential elections. At its peak in early 1984, the Diretas Já! claimed 1.5 million participants. While the PMDB led Diretas Já! at the institutional level, the PT led on the street (Secco 2011: 113). In November 1983, the PT filled São Paulo s gigantic Pacaembu Stadium for the first Diretas Já mass protest. An early petista observed that no other party could have begun to mobilize such a massive crowd (interview with Alli). The PT s street-level leadership during Diretas Já! enhanced the party s status in civil society, as did the PT s decision to boycott the indirect election of 1985 (about which more below). Secco (2011) writes that due to the PT s boycott, its number of supporters and presence in social movements grew (e.g., in student movements) (Secco 2011: 119). After the amendment for direct elections (the Dante de Oliveira Amendment) narrowly lost in a congressional vote, the PMDB accepted the result and nominated a leader, Tancredo Neves, to run in the indirect presidential election of The PT deliberated whether to send its federal deputies to the Electoral College to vote for Neves. In order to reach a decision, the PT held an internal vote. Twenty-thousand petistas participated, and an overwhelming majority, eighty-six percent, voted to boycott the Electoral College. 203 Interview with Hamilton Pereira in de Moraes and Fortes (eds., 2008: 266). 204 In 1983, the Brazilian economy ground to a halt, and annual inflation reached nearly 250 percent. 137

146 Months before the mayoral elections of 1985, the PT s second electoral cycle, many [PT leaders] expressed doubt that the party would last out the year (Keck 1992: 235). The failure of the Diretas Já movement, as well as internal conflicts over the decision to reject a compromise solution, had produced yet another internal crisis (Keck 1992: 154). The PT s perceived inability to translate civil society strength into institutional efficacy was causing a full-scale crisis of identity in the party (ibid). The mayoral elections were widely seen as the great test of the party s organizational viability. Unless the results were a vast improvement over the party s performance in 1982, party activists would more than likely opt to concentrate their energies entirely on movement organizations, not the PT (ibid). Yet the PT performed unexpectedly well. The party had mounted campaigns in nearly every state capital, and petista candidates registered breakout performances across the board. The PT s Maria Luiza Fontanelle won the mayoralty of Fortaleza (Brazil s fifth largest city), and unlike in 1982, PT candidates finished second or third in a number of other major municipal contests. In Vitória, Espírito Santo and Aracajú, Sergipe, for example, the PT placed second, and in the all-important São Paulo mayoral race, the PT s Eduardo Suplicy finished a close third with twenty percent of the vote. In part, the PT s performance simply reflected the party s new civil society ties, developed since Crucially, however, the PT had also begun to solidify a distinctive party brand. Although the PT s firm stand against indirect election led to the party s short-term marginalization in the spheres of institutional politics and the national media, it helped cement the party s image as a principled political actor, less willing to compromise for short-term political gain than the center-left, establishment (situacionista) PMDB. Secco (2011) writes that this choice [to boycott the Electoral College]...reinforced [the PT s] external identity as a group of social opposition (119). The PT benefited, electorally, from its enhanced stature among pro-democracy advocates, receiving a large portion of the protest vote. 138

147 After 1985, the PT suffered major electoral disappointments (e.g., Lula s losses in 1989 and 1994) but never ran a serious risk of electoral collapse, as in 1982 or The party made steady electoral progress for the next decade. In the 1986 congressional elections, the PT doubled its congressional contingent, jumping from eight to sixteen federal deputies. The newly elected deputies included Lula, Olívio Dutra, and several other major party leaders. Former Marxist and PT leader, José Dirceu, won a seat on the São Paulo state assembly. 205 The 1988 elections were a great leap forward. The party elected thirty-one mayors, up from one in 1985 and Luiza Erundina won the PT s major executive post, the São Paulo mayoralty. The PT won three state capitals (SP, Porto, Vitória), six mid-sized industrial cities in SP state (including three in the ABCD region), 206 and three small cities in Minas Gerais (Hunter 2010: 81). The PT also performed well in rural districts where the party had worked closely with the landless movement and/or rural unions, labor organizations (Keck 1992: 157). After 1988, the PT, and especially Lula, became perennial contenders in major national and subnational elections. Lula came within a hair of winning the 1989 presidential contest, 207 and placed second again in 1994 (about which more below). In 1990, the PT doubled its legislative contingent a second time, electing thirty-five federal deputies and winning over ten percent (10.2) of Chamber vote share, and in 1992, the PT, despite losing control of the São Paulo mayoralty, elected fifty-four mayors (up from thirty-six in 1988) and won its first governorships: the Federal District (Brasília) and Espírito Santo. 205 Thus, after four years dedicated to a base, the PT s top national leaders in Keck s (1992) words, the party s best-known and most legitimate spokespersons (224) was returning to institutional politics. 206 Diadêma, Santo André, São Bernardo do Campo 207 The race, in a commonly used phrase, pitted unorganized Brazil (Collor) against organized Brazil (Lula). Lula won in the cities, with the votes of organized workers and sympathetic middle-class progressives, but he fared extremely poorly among the unorganized poor and working class, especially in the undeveloped Northeast, where Collor and allied conservative patronage and media machines dominated. Coronelismo, both material and electronic, narrowly won the election for Collor. 139

148 Chapter 3 Avoiding schism in the early PT: the indispensability of Lula da Silva So far, the chapter has examined the causes and consequences of the early PT s organizational strength. Low access to resources and media created an electoral incentive for the PT to build a strong organization, while access to civil society union, church, and Marxist structures gave the PT the capacity to succeed. Due primarily to these conditions, the early PT developed a large base of activists and members, organized locally, with a high level of commitment to the party s aims. Organizational strength accounted for the early PT s steady electoral development, and the party s grassroots capacity and ethos fortified it amid early electoral crisis. The previous section identified the PT s perceived popular essence as the main collective incentive fueling early petistas. Members identification with and commitment to the PT s popular character and narrative, however, did not eliminate centrifugal forces within the party. Indeed, the sincere ideological commitments of petistas, while useful for elongating activist time horizons and preventing electoral collapse, arguably made the party base less pragmatic and thus threatened cohesion. As a rule, sectarianism plagues the most ideological parties and movements (e.g., Marxism), given that internal currents, despite differing from each other in seemingly minor ways, often prioritize ideological purity over political pragmatism and compromise. Moreover, PT blocs and factions differed in major, not minor, ways. From inception, the PT was furiously expanding, first, to achieve official status and, second, to prepare for the 1982 elections. In this process, the party prioritized territorial reach over internal coherence. Leaders wanted to incorporate as many societal actors as possible, from across the left spectrum. In order to attract the maximum range and number of feeder groups, the party cúpula openly welcomed and embraced internal difference and, in order not to alienate 140

149 any important actors on the radical or moderate left, refrained from espousing a specific ideology. In the words of one Marxist ex-petista, if party leaders had officially endorsed any systematic analysis of the nation s current ills and proper direction, they would have alienated important groups. 208 The vast majority of early PT leaders, new unionists included, endorsed socialism in a broad sense. Yet party leaders faced a difficult balancing act. On the one hand, they rejected rigid Marxist doctrine, which just a fraction of the leadership and base espoused. On the other, they sought to attract Marxists by making the PT s official ideology unspecific, open-ended, and hence contestable. A vague petista socialism, or democratic socialism, resulted. PT leaders, including Lula, openly identified themselves and the party as socialist, but they defined their socialism primarily in negative terms, as a rejection of bureaucratic socialism and social democracy (Ribeiro 2010: 214). 209 The more positive formulations of petista socialism possessed little specific content: the emancipation of workers by workers, to be realized in practice through internal and external democracy, not worked out a priori. 210 Lula articulated o socialismo petista in his address at the PT s first Encontro Nacional in The PT succeeded in attracting a very heterogeneous set of actors, who not only differed in ideology, but also hailed from disparate regions, belonged to different socioeconomic strata, and identified with different subcultures. PT members and factions spanned unreformed Marxism and the moderate left, the popular and middle sectors, the private and public sectors, the full range of education levels, and developed and undeveloped areas of the coun- 208 Interview no Defined strictly in negative terms as a rejection of bureaucratic socialism and social democracy the PT s democratic socialism never advanced, in almost thirty years, beyond Lula s 1981 speech (Ribeiro 2010: 214). 210 Keck 1992:

150 try. 211 Secco (2011) emphasizes the enormous regional diversity of the early PT, observing that local PT networks varied widely from state to state, and even municipality to municipality, in their civil societal and ideological profiles. 212 Echoing the analysis of Keck (1992), he stresses that the early PT was not just one thing, but instead a patchwork of local organizations and subcultures, the character of each local branch and activist network largely determined by the identity of the leading feeder group/s in the area. 213 Although the PT s internal heterogeneity produced a range of conflicts and tensions, differences between PT moderates (primarily new unionists) and radicals (overwhelmingly Marxists) posed the greatest challenge to cohesion. The early PT housed a large number of loosely organized new unionists and left Catholics and a large number of small but very well-organized Marxist parties and organizations. From inception, the PT, in order to attract Marxists, recognized the direito de tendência, which permitted the existence, within the PT, of formally organized factions (tendências) that possessed specific ideological profiles and sought to influence the PT s programmatic direction and occupy both publicsector and party positions. The party thus originated as a party of, not with, tendências. 214 Tendências, Ribeiro (2010) writes, are part of the petista DNA (186). Crucially, the PT leadership also tacitly permitted dupla militância, enforcing virtually no restrictions on tendências activities outside the party. Tendências were allowed, although not encouraged, to remain separate parties and explicitly to treat the PT as a tactical party in the service of separate, revolutionary ends. Most did. On a standard view within most 211 In the words of one founding petista, previously a founder of the Revolutionary Brazilian Communist Party (PCBR), you had a large number of MDB legislators, you had the university movement, the art world, the world of journalism, the public sector you had many areas that came together to create the party (interview with Apolônio de Carvalho in de Moraes and Fortes (eds., 2008: 22). 212 Talk by Lincoln Secco at PT Diretório Zonal in Pinheiros, São Paulo city. 213 Talk by Lincoln Secco at PT Diretório Zonal in Pinheiros, São Paulo city. 214 Phrase taken from D Amato (1964). 142

151 Marxist tendências, new unionists lacked full class consciousness and, barring conversion, would never pursue revolutionary ends. These tendências thus treated the PT, either explicitly or implicitly, as an electoral front and regarded their own participation in the PT as a tactic, not a strategy. Indeed, new unionists did not subscribe to Marxism, at least its more revolutionary tenets. According to one Marxist PT founder, new unionists feared Marxists revolutionary discourse. 215 In one characteristic reaction to radical members of the PT in Pará, a rural union leader described the discussions surrounding the formation of the CUT in 1983: There was a part [of the text synthesizing the discussions] that talked about getting into power. The language was very... [trails off ] I think they were people linked to the MR Workers discomfort with Marxists resulted not only from ideological differences, but also from interpersonal barriers rooted in class difference. New unionists political views had arisen, primarily, from working-class and union experience, not from reading and education. PT workers often seemed not to respect PT Marxists, viewing them as disconnected intellectuals, materially comfortable in their private lives, and limited in their experience of genuine struggle. In early 1981, an internal PT document, circulated to clandestine Marxist parties associated with the party, tellingly read: If you want to enter the party that the working class, the actual working class, is organizing, fine. You will be welcomed. But don t come as owners of the truth, dictators of the rules for the masses (Teixeira 1981, cited in Secco 2011: 95; emphasis not in the original). Yet PT workers also, at times, seemed insecure about, and even resentful of, Marxists higher education levels. According to one Marxist PT founder, new unionists knew nothing about political theories, philosophies, and Lula, initially, [was] so conscious of his origins that he 215 Interview with Apolônio de Carvalho, de Moraes and Fortes (eds., 2008). 216 Interview with Avelino Ganzer in de Moraes and Fortes (eds., 2008: 171). 143

152 didn t want to talk to intellectuals and students. 217 A Marxist ex-petista noted that, early in the party s history, Lula openly admitted to having no idea what Marx or Lenin argued. 218 For all these reasons,...there was, in the PT, a kind of base-level, anti-intellectual worker culture, even though [the party] had attracted many intellectuals since its founding (Secco 2011: 94). A former urban guerrilla and initial PT enthusiast recalled that at an early PT meeting in Rio de Janeiro, the new unionists in control of the meeting did not allow Marxist intellectuals to take the microphone, only allowing operários to speak. 219 Critiques of the radical tendências by moderate petistas, Ribeiro (2010) writes, are as old as the tendências themselves (186). From the party s early days, such critiques crop up regularly in internal party communications and usually center on allegations of duplicity, insularity, and rigid dogmatism. 220 Interestingly, new unionists encouraged the formation of núcleos de base, in part, to create union- and church-based counterweights to the Marxist tendências (Secco 2011: 78). * * * In sum, given the internal weight of ideologues in the PT as a whole, and the serious ideological and sociological differences between PT groups, schism posed a threat. In the words of Hamilton Pereira, Marxist PT leader, Initially, [the PT was] very strong and very fragile. It [was] strong because it [was] the meeting place between the [Marxist] left...and a nascent workers movement. Now, [it was] very fragile because it was...born from a break with both the III International and social democracy... Only the PT [was] more complex than that. It incorporate[d] people with a Church background and the unionists... What was the result of this? The threat of low cohesion, of fragmentation (da falta de coesão, da fragmentação) Interview with Apolônio de Carvalho, de Moraes and Fortes (eds., 2008):. 218 Interview no Interview with Estevão. 220 Ribeiro (2010): 332, footnote 8; Dutra Interview with Hamilton Pereira in de Moraes and Fortes (eds., 2008: 264). 144

153 Indeed, on numerous occasions during the formative phase, the PT seemed on the verge of schism. Keck (1992) summarizes that in the first half of the 1980s...it often appeared that the difficulties in resolving internal and environmental challenges would destroy the party (18). In particular, throughout the 1980s and first half of the 1990s, conflicts between moderate PT leaders and radical elements of the PT base threatened the party s integrity. The remainder of the section describes the PT s central internal conflicts and analyzes why the PT avoided fatal schism during the formative phase. The PT s luta interna From the beginning, petistas resolved their differences, primarily, in a formal internal arena, the sistema político petista. In 1980, the PT established a formal internal democracy in which the rank-and-file could assert control, directly and indirectly, over internal dialogue and governance (Keck 1992: 91). Drawing on aspects of both new unions and CEBs, the PT, from the beginning, put a premium on participatory democracy: The PT s organizational proposta was based, essentially, on the principle of participatory democracy (Secco 2011: 81). While the PT s flexible ideology would foster a sense among the PT s internal groups, including the more radical minority groups, that they could influence the party s direction, the party s formal and informal participatory democratic institutions would provide channels for the bottom-up shaping of party policy and principle. 222 The formal rules of the sistema político petista would go virtually unchanged until These rules combined proportional representation (with a ten-percent minimum) for forming the national and subnational offices, or Diretórios, and winner-takes-all majoritarianism for forming executive committees, or Executivas. All PT members were eligible to participate in official party meetings, Encontros, at the municipal or (in large cities) zonal level. Municipal 222 Interview no. 3 and communication with Ribeiro. 145

154 and zonal Encontros sent delegates to state Encontros, which in turn sent representatives to the national Encontro. At every level, Encontro attendees elected representatives to the national, state, municipal, and zonal offices, or Diretórios, by proportional representation. Tendéncias had to receive ten percent of the Encontro vote in order to receive a minimum of representation in the corresponding Diretório. Encontro members typically voted as part of tendências, the parties of the petista political system (Ribeiro 2010: 186). The tendência or coalition of tendências that held the absolute majority in a Diretório appointed all members of the corresponding executive committee, or Executiva. Only in Rio Grande do Sul did the state Diretório allot internal executive posts via proportional representation. 223 A large number of Marxist feeder organizations became tendências immediately upon joining, including the particularly effective and influential Trotskyist groups, (Causa Operária, Convergência Socialista, O Trabalho). Given the number and efficacy of Marxist tendências, the tendências, in informal PT parlance, became shorthand for the revolutionary leftist currents and parties acting within the PT (Keck 1992: 118-9). During the period, Marxist tendências exerted enormous influence internally, both in organizing the party at the local level and in shaping the party program. As already noted, in the initial rush to achieve registry and expand the ground organization in preparation for the 1982 elections, the national leadership devoted scarce energy to issues of programmatic coherence and internal discipline, partly due to lack of time (Keck 1992: 121), partly in order to avoid imposing on segments of the base. This state of affairs created fertile terrain for well-organized Marxist petistas who wished to control the party message and program in their local bastions. 223 Interview with Raul Pont in de Moraes and Fortes (eds., 2008: 225). 146

155 On the PT s official analysis, the incoherence and, at times, the militant stridency of the party s November 1982 campaign harmed the party electorally. After the electoral debacle, the internal influence of Marxists became a top concern for the PT s moderate, largely new unionist leadership. Despite their numbers, moderate PT leaders lacked the discipline, and internal focus/efficacy of the PT s left tendências. Party leaders concluded that the interaction of Marxist efficacy and organizational decentralization, and the resulting absence of a single and minimally pragmatic party line, had thwarted and would continue to thwart the party s external electoral ambitions. By underinvesting in the luta interna, moderates had allowed radicals undue influence over the party message, ultimately handicapping the party at the ballot box. Moving forward, electoral progress would require a minimum of coherence and pragmatism in the party program. The 1982 election marked a watershed in PT history for two reasons. 224 First, PT moderates created a large, formal, centrist bloc, the Articulação dos 113 (henceforth Articulação). In creating the Articulação, the PT s moderate center set out to formalize and consolidate its dominance, with a view toward imposing at least a modicum of programmatic moderation on the entire party. On Keck s (1992) summary analysis, the founders of Articulação sought to create a unifying political center (Secco 2011: 93) and impose a relatively unified vision of the party s nature and goals (Keck 1992: 114). The formation of Articulação was an attempt to consolidate leadership of the party... It represented an effort to impose a relatively unified vision of the party s nature and goals, not to the point of eliminating factional differences, but at least as the expression of a clear majority (Keck 1992: 114). Rodríguez (n.d.) similarly characterizes the Articulação as an attempt by Lula and allies to formalize their status as the PT s dominant coalition. Ricardo Kotscho simply called the 224 The electoral setback of 1982 proved crucial in [the PT s] early development (Keck 1992: 124), leading to a new period of PT history according to some. From talk by Lincoln Secco at PT Diretório Zonal in Pinheiros, São Paulo city. 147

156 Articulação the center of the PT. 225 Second, as discussed earlier, during the same period (late 1982 to early 1983), PT labor leaders, many simultaneously involved in the creation of Articulação, returned to the base on the view that the party had grown too distant from its civil society roots. As a central element of the return to the base, PT labor leaders created and nurtured the CUT. PT moderates hoped and intended, with the CUT initiative, to increase the internal leverage of moderates by strengthening the new unionism. Thus, both the creation of the Articulação and the return to the base were attempts to consolidate [leadership] of the party in the hands of labor movement leaders and those they considered their allies (Keck 1992: 153). The rise of Articulação Articulação s founding leaders spanned new unionists, left Catholics, and Marxist leaders and intellectuals. Lula did not formally join Articulação but, according to Secco (2011), led the bloc symbolically (93). Articulação s leaders included all the strongest personalities 226 and main names of the period (Ribeiro 2010: 187), including Lula (informally), Olívio Dutra, and Benedita da Silva. According to Marxist petista Hamilton Pereira, radicals who joined the Articulação did so as a service to the consolidation of the PT (F and F 2008: 265). In June of 1983, 113 party leaders published the Articulação s manifesto, the Manifesto dos 113. In positive terms, the manifesto continued in the tradition of petista socialism. Articulação affirmed socialism but kept its positive vision vague and open-ended, to be contested and realized in practice through mass participation and democracy, both in and out of the PT. In the manifesto, the Articulação also situated itself between the vanguardist and elitist 225 Interview with Kotscho. 226 Interview with Donato. 148

157 poles of the party on the one hand, the Marxist tendências and, on the other, more pragmatic, conservative sectors that placed too much emphasis on short-term electoral politics and insufficient emphasis on the party s civil society linkages (Keck 1992: 114; Secco 2011: 94). In truth, however, the goal of reining in the Marxists constituted the primary motivation for creating Articulação. While affirming socialism, the Manifesto dos 113 defined Articulação in clear opposition to traditional Marxism. Various tenets and characteristics of Marxism received censure, including bureaucratization, vanguardism, the treatment of the PT as a tactical front, and the treatment of the present conjuntura as an early, bourgeois-democratic stage in Brazil s development toward communism (Secco 2011: 94; Ribeiro 2010: 214). The manifesto reserved its most pointed criticisms for petistas de duas camisas who subordinate themselves to parallel commands and enclose themselves in the proposal of a traditional vanguardist party that deems itself representative of the working class. 227 Ribeiro (2010) thus summarizes that PT leaders created the Articulação in order to maintain hegemony over the leftist factions in particular: The construction of a large faction [the Articulação] was the measure found to attempt to restrain the activity of the leftist groups (186, 187). Along similar lines, Secco (2011) observes that the Articulação arose more as an antitendência, an effort to contain (enquadrar) the tendências and centralize the party (Secco 2011: 123). The Articulação quickly established a dominant internal position. In the Articulação s first internal election in 1984, bloc members won two thirds of the delegate seats to the third National Encontro in From 1984 to 1986, the Articulação began to impose on the entire party its critiques of the other tendências behavior (Ribeiro 2010: 188). 228 Among 227 Documento dos 113. Articulaçã dos 113, São Paulo, June Interestingly, the Articulação s first major imposition targeted the more elitist, institutionalist sectors of 149

158 other things, Articulação leaders insisted on a more inclusive and less strident campaign for the 1985 mayoral elections. Keck (1992: 155) thus observes that in 1985, the PT adopted a more inclusive and humorous campaign tone and even made better use of media in certain elections (e.g., the São Paulo mayoral race). The party ran more middle-sector candidates and paid more attention to the middle sectors and other demographics (e.g., housewives) (155). After the PT s unexpectedly strong performance in the 1985 elections, the Articulação consolidated its dominance. Basking in the glow of the PT s strong electoral results, Articulação wrested or retained control of the vast majority of zonal, municipal, and state Diretórios and increased its dominance within the Diretório Nacional, having elected nearly three quarters of the delegates at the IV Encontro Nacional. Executive proportionality and dupla militância As the Articulação consolidated its dominance, tensions between PT moderates and radicals escalated. The creation and consolidation of the Articulação both exacerbated and brought into relief the internal marginalization of the Marxists. Radicals began to chafe at the rules of the petista political system, which patently benefited moderates at the expense of radicals. As described earlier, PT formally practiced winner-takes-all majoritarianism in the allotment of executive posts at all levels. With majorities in Diretórios at the federal, state, municipal, and zonal levels, the Articulação, after 1984 and 1986 Encontros, monopolized the PT s executive functions, except where subnational Executivas informally practiced the party. In 1985, the PT membership internally voted in favor of boycotting the indirect elections. The no vote prevailed, in part, because the left tendências most opposed to the indirect elections mobilized their members effectively. Following the vote, the PT s national Diretório, controlled by the Articulação, mandated that all federal deputies abstain. When three (of sixteen) PT federal deputies Airton Soares (SP), Bete Mendes (SP), José Eudes Freitas (RJ) participated and cast votes for the PMDB s Tancredo Neves, the DN promptly expelled them from the party. 150

159 proportionality. 229 Absent the formal implementation of proportionality in the CEN, the Articulação would monopolize the PT most important executive body indefinitely. The majoritarian system for the Executivas was one of the main bones of contention between the Articulação and the leftist factions throughout the 1980s (Ribeiro 2010: 188). By the May 1986 Encontro, the PT s left tendências feared irrelevance and, increasingly, began to openly question whether they should remain in the PT. For the Articulação, unity of action remained at the top of the internal agenda. Despite internal differences, the PT needed to project an image of unity and a coherent message. PT radicals, in the view of Articulação leaders, continued to undermine this imperative. Moderate exasperation with left tendências entrismo ( betweenism ) and abuse of the Direito de Tendência (factional rights) intensified around the time of the IV Encontro Nacional (May 1986). Only a month earlier, in April, authorities had arrested a group of radical petistas for robbing a Banco do Brasil branch in Salvador (Ribeiro 2010: 188). At the Encontro, the PT expelled the implicated petistas and formally resolved to regulate (regulamentar) and contain (enquadrar) the tendências in the next Encontro Nacional. The question of proportionality [in the Executivas] could only be discussed after that enquadramento. In the meantime, the majoritarian criterion would remain in effect (Ribeiro 2010: 188). In May of 1987, O Trabalho, one of the PT s most important Trotskyist factions, split in two, with most joining the Articulação and the rest remaining a separate tendência (called the Agrupamento). Throughout 1987, left petistas forcefully demanded proportional representation in the CEN, calling the majoritarian alternative nakedly anti-democratic. In an April 1987 pamphlet, 229 In practice, some state, municipal, and zonal Diretórios used proportionality in the allotment of executive posts in order to prevent the defection of minority groups. In the all-important National Executive Committee (CEN), however, the Articulação maintained absolute control. 151

160 O PT as Tendências e a Luta Interna ( The PT, the Tendências and Internal Struggle ), petista Augusto de Franco wrote: Mere obedience of statutes by individuals...and identification with the platform are not sufficient to normalize democratic relations between the diverse tendências within the PT and between the tendências and party organs. Granted, everyone must endorse the majority position after internal deliberations...so that for the masses, there [is] only the PT, and not its currents. Yet the PT s internal political system must enable the various tendências to defend their stances within the party. The pamphlet centered on the need for proportional representation in all party organs: Diretórios and Executivas. Later, the pamphlet reiterated the necessity of correcting deformities in our internal democracy (guaranteeing, for example, proportional participation of the minorities in the Executivas ). 230 In a June 1987 pamphlet, Sobre a construção partidária ( On party-building ), Marxist PT leader Raul Pont extolled proportional representation in the Diretórios but stressed: We need to consecrate that at the level of the Executivas as well, incorporating...this rule explicitly and clearly so that it does not depend on the good will of the...majorities. 231 In December 1987, the PT held its fifth Encontro Nacional, standardly considered the most important in the history of the PT (Secco 2011: 121). The Articulação won sixty percent of the delegate vote a wide majority, but not as wide as in At the Encontro, left tendências, for the first time, formally proposed the implementation of proportionality in the CEN. The Articulação, despite some defections, succeeded in blocking the motion. In a formal resolution on the issue, the Articulação acknowledged that, at the subnational level, PT Diretórios already practiced proportionality on an informal basis but cited the need for greater cohesion at the top. The Articulação resolved that the National Diretório would only approve proportionality in the CEN if radical tendências first consented to the abolition of 230 From Edgar Leuenroth archive. 231 From Edgar Leuenroth archive. 152

161 dupla militância. The resolution affirmed that all PT tendências must treat the PT as a strategic vehicle, not as a tactical front in the service of separate, fundamental objectives. Handing over executive functions to tendências with separate party projects, the resolution argued, would prevent unified action and lead to decision-making paralysis. 232 The general resolution passed, but due to internal negotiations with the the most problematic tendências as well as the demands of the 1989 presidential campaign, the PT did not regulate the tendências or institute proportionality across the board until 1990 (Ribeiro 2010: 189). In April 1990, the DN abolished the right of tendências to claim independent status as parties, subordinate members to internal decisions, subordinate themselves to international bodies, engage in clandestine activities, disseminate independent programs, man independent locales, and more (Ribeiro 2010: 189). In May of 1990, the DN instituted proportionality for executive committees at all levels, including the CEN. At its first National Congress in 1991, the PT lowered barriers to entry for representation in the Diretórios and hence the Executivas by abolishing the ten-percent minimum. In addition, the PT took several measures to empower base-level petistas without strong ties to a particular tendência, and hence to limit the decision-making influence of small, well-organized left tendências (e.g., the institution of a direct method i.e., closed primaries for executive candidate selection). 233 The implementation of proportionality in the CEN observably strengthened left tendências commitment to the PT. The vast majority of left tendências accepted the internal regulatory reforms of period. Some of these, including the PRC and PCBR, dissolved and either joined preexisting tendências or created new ones. Others continued as before, but in deference to the new regulations (Ribeiro 2010: 192-3). As of mid-1990, PT tendências 232 Discussed in Ribeiro 2010: PT: O desafio de ser partido, from the offices of PT Deputies Eduardo Jorge and José Genoino (1992). From Edgar Leuenroth archive. 153

162 numbered ten in total: the Articulação and nine smaller Marxist tendências. 234 In 1992, the PT expelled two Trotskyist tendências, Convergência Socialist and Causa Operária, for violating the newly imposed restrictions on tendências. The Marxist surge and risk of schism The collapse of socialist regimes in the Soviet Union and Eastern Bloc produced a realignment of forces within the PT. First, the crisis of international socialism led some Marxist factions to renovate. The PRC, a Marxist tendência with reasonable internal weight led by José Genoino, renounced Marxism at the VII Encontro in 1990 and moved definitively to the PT s right wing, alongside the more moderate sectors of Articulação. More importantly, the collapse of socialism created a fissure within Articulação that ultimately led to the group s schism. Whereas Articulação s moderate majority saw the collapse of the USSR and Eastern European regimes in a positive light, the Marxist and ex-marxist wings of Articulação, like the PT s left tendências, refused to expressly condemn the socialist experiments that were collapsing (Ribeiro 2010: 197). The left dissidents within Articulação claimed that the bloc s moderate wings, in pursuit of electoral success, were abandoning the PT s commitment to socialism and, in their excessive focus on institutional politics, distancing the party from civil society, its historical source of strength (Melo 1994: 86-9; Ribeiro 2010: , 203). Articulação held a majority at the VII Encontro, and the formal resolution approved at the Encontro, entitled O Socialismo petista, reflected the tendência s ideological divisions. In keeping with tradition, the text affirmed socialism and firmly rejected social democracy, both as an end in itself and a means to socialism (Ribeiro 2010: 199). The resolution shied away from a specific, positive vision, Marxist or otherwise. As before, petista socialism 234 CS, DS, FS, Luta pelo Socialismo, TM, OT, VS, Vol Proletária, Nova Esquerda (Ribeiro 2010: 190-1). 154

163 would be built from practical experience (ibid). Yet in a demonstration of left dissidents influence, the text also warned of a capitalist regression in the Eastern Bloc countries (ibid). Equally, the resolution reflected efforts by Articulação s moderate leaders (e.g., Lula, Dirceu) to bridge the tendência s growing divide. The text s formulators endeavored, when possible, either to synthesize or navigate between the tendência s contrasting, even contradictory, left and right wings. They resolved, for example, that the PT should both lead (ser dirigentes de) and dialogue with (ter interlocução com) civil society, not one or the other. By the time of the PT s polarized I Congresso (1991), however, conflict between moderates and radicals had become significantly sharper (Ribeiro 2010: 197), and internal conflict had begun to hamper the Articulação s internal electoral fortunes. Articulação elected a plurality, but not a majority, of delegates to the congress. Over the next two years, conjunctural debates regarding Collor s impeachment, the presidentialism/parliamentarism plebiscite, and Luiza Erundina s controversial term in São Paulo further divided Articulação (Ribeiro 2010: 200-3). Meanwhile, external critiques of Articulação from the PT s Marxist tendências were resonating with and galvanizing large segments of the party base. In 1992 and 1993, Marxist tendências won majorities in a series of major municipal Diretórios, including those of São Paulo, Rio de Janeiro, Salvador, Fortaleza, Porto Alegre, Vitória, and Belém (Hunter 2010: 120, note 18). These municipal results affected the composition of state Diretórios, which in turn would affect the composition of the Diretório Nacional. Articulação formally split in Prior to that year s VIII Encontro, Articulação s left dissidents had released a manifesto, A Hora de Verdade, outlining their critiques of Articulação s moderate wings. At the Encontro, the Articulação s left dissidents acted as an informal bloc, temporarily called Hora de Verdade (HV), and allied with the left tendência, Democracia Socialista (DS), combining their teses into a single tese, Opção de Esquerda 155

164 ( Left Option ). 235 The core of the Articulação, under the new name Unidade de Luta (Hunter 2010: 125, note 33), published a separate, individual tese, as did Democracia Radical ( definitively situated on the PT s right wing (Ribeiro 2010: 204)) and a coalition of the PT s most extreme left tendências. The HV/DS tese won majority support (fifty-six percent), giving PT radicals their first national majority. The text approved at the VIII Encontro Nacional unambiguously endorsed socialism, embraced revolutionary struggle, and rejected the third way between socialism and social democracy. The text also exhorted a second return to the base. 236 Months after the Encontro, the Articulação s left dissidents would register as a formal tendência under the name Articulação de Esquerda (AE). The PT s official platform and preliminary campaign strategies for the 1994 presidential election reflected the internal ascendance of the Marxist tendências. The platform evoked radical change brought about by state action, promising transformation and even revolution. With regard to campaign strategy, left tendências insisted on highlighting Lula s status as an operário (laborer) and on allying exclusively with left-wing, predominantly socialist parties such as the PCdoB, the Brazilian Socialist Party (PSB), the United Socialist Workers Party (PSTU), the Popular Socialist Party, and the Green Party (PV). 237 The platform did depart however, from the strident, unabashed socialism of 1989, demonstrating the evolution and continued influence of the moderate minority. 238 As the 1994 election neared, moderate/radical tensions escalated. In some respects, these 235 Prior to 1986, the DS had been an independent party as well as a powerful left tendência in the PT. In 1986, the DS dissolved and became a PT tendência only. In subsequent years, the DS tacked toward the center of the PT political system, becoming part of the PT s moderate left (Ribeiro 2010: 204). 236 Hunter (2010): Hunter 2010: Hunter (2010) summarizes, [t]he overarching difference between the 1989 and 1994 programs was that the latter backed away from professing socialism as the ultimate goal (118-9). 156

165 tensions mirrored, at the national level, internal PT struggles that had previously occurred at the subnational level. Indeed, since the mid-1980s, the most intense conflicts within the PT had occurred when PT executive candidates and elected PT executives, confronted with the realities of campaigning and governing on a large scale, sought to impose pragmatic measures and confronted resistance from internal PT elements. The first major episodes occurred in large municipalities, including Fortaleza (1986-9), Campinas ( ), 239 and São Paulo ( ). In São Paulo, prefeita Luiza Erundina (PT) had sought unsuccessfully to sell a pragmatic, pluralistic governing approach, the PT de resultados, on the local and national PT. 240 Erundina s struggles with the PT organization had continued without resolution for four years, hampered her administration, and played a significant role in the PT s defeat in São Paulo in 1992 (Hunter 2010: 89-92). 241 Ironically, in the lead-up to the 1994 presidential election, Lula and his closest associates (e.g., José Dirceu), who had weighed in against Erundina just a couple of years earlier, began to fear for the PT s chances in the upcoming presidential election and moved to impose a more pragmatic campaigning strategy and more moderate platform on the Marxist-controlled Diretório Nacional. 242 Importantly, however, what Luiza Erundina failed to achieve in São Paulo, Lula succeeded in achieving nationally. After the formation of the extreme left Diretório Nacional in 1993 (Hunter 2010: 121), Lula ran for party president and won overwhelmingly. Lula used his mandate to inject a dose of pragmatism into the party s 239 In February 1991, new unionist Jacó Bittar, petista mayor of Campinas, was expelled from party after developing strong links with the [clientelistic] Quercia machine and stating that he was no longer bound by decisions of the Campinas diretório (Keck 1992: 235). 240 The Erundina administration favored social policies that would benefit all disadvantaged sectors of society, including the vast informal poor, and not just PT constituents (e.g., formal-sector workers). The municipal and national PT and CUT forcefully resisted, demanding the union-focused agenda that Erundina promised in the campaign. Erundina also awarded all major cabinet positions to allied factions, alienating important sectors of the São Paulo Diretório (Hunter 2010: 91). 241 Secco (2011) writes that the PT was its own greatest enemy during Erundina s term ( ) (130). 242 Hunter (2010):

166 1994 presidential bid. Shortly after winning, in late May of 1993, Lula demanded that the PT s xiitas (Shiites, or radicals) give him a margin of autonomy to run his campaign and threatened to call off his candidacy if the trend toward radicalization continued (Hunter 2010: 121). Subsequently, Lula met quietly with business elites, contacted centrist party leaders about possible alliances, insisted on moderating the PT s privatization and expropriation platforms, and successfully advocated the vice presidential nomination of businessand military-linked Aloizio Mercadante (Hunter 2010: 121-2). Lula managed to impose these measures without incident. 243 Despite his efforts, Lula lost decisively to the PSDB s Fernando Henrique Cardoso (FHC). Save the PT s 1982 electoral debacle, this loss proved the most devastating and challenging in the party s history. With good reason, petistas had entered the campaign confident of victory and, on the prevailing internal analysis, lost unnecessarily. FHC had played a leading role in devising and promoting the anti-inflationary Real Plan as finance minister. The Real Plan enjoyed immense public support, having stabilizing Brazilian price levels after years of high inflation and periodic hyperinflation. Imprudently, moderates and radicals alike had chosen to oppose the Plano Real. While Hora de Verdade and other left tendências had publicly demeaned price stability as a bourgeois concern, moderate factions [had] charged that stabilization had been accomplished by squeezing salaries, and members of the Lula camp [had] openly predicted the ultimate failure of the plan (Hunter 2010: 125). Lula s pragmatism notwithstanding, PT moderates also believed that the party s continued association with leftism, in voters minds, had contributed to the 1994 presidential loss. Moving forward, they concluded, the party would not win national power on an uncompromising left platform. This conclusion marked an important shift. Prior to the 1994 election, many 243 On Hunter s (2010) summary, [a]s radical as the program, tone, and tactics [alliances, caravans] of the 1994 campaign were, without the moderating influence of Lula and his associates they would have been markedly more so (Hunter 2010: 120). 158

167 PT moderates believed that a solidly left party could win national power in Brazil. This view stemmed, in large measure, from Lula s strong performance in PT moderates and radicals alike originally took Lula s 1989 performance as evidence that left-wing appeals resonated with the median voter in Brazil. After 1994, a revisionist view of the 1989 election took hold among PT moderates. Leftism as such had not benefited Lula. Instead, the special circumstances of [the 1989] campaign a highly polarized contest between a populist right-wing candidate associated with the military regime and a candidate who represented a break with that style and era accounted for Lula s strong showing (Hunter 2010: 126). After Lula s 1994 loss, PT moderates concluded that future presidential victory would require a centrist ideological shift, the provision of concrete immediate material benefits, a deemphasis on program, and an increased emphasis on Lula s charismatic personality (Hunter 2010: 126). Moderates thus stepped up their efforts to defeat radicals internally. The PT s X Encontro Nacional, the most polarized in the PT s history, took place in The Articulação, in alliance with the DR, narrowly won after picking off, with offers of top party positions, two AE elites who had played important roles in the AE s creation (Cândido Vaccarezza, Rui Falcão). The contest for party president pitted José Dirceu of Articulação, whom Lula explicitly supported, against Hamilton Pereira of AE. During pre-vote speeches, César Benjamin of the AE made an unusual frontal attack on Dirceu, to Dirceu s surprise. 244 On Hamilton Pereira s telling, Benjamin emphasized that Dirceu had received money from businesspeople, and that it was absurd for a workers party to rely on money from big business... In sum, he said everything that his liver dictated. 245 Dirceu edged out Pereira by two to three 244 Interview with Hamilton Pereira in de Moraes and Fortes (eds., 2008: 273). 245 Interview with Hamilton Pereira in de Moraes and Fortes (eds., 2008: 273). 159

168 percentage points, and PT centrists once again assumed control of the DN and CEN. In a vivid demonstration of the party s deep fissures, however, the PT did not manage to form the full CEN until January of 1997, due to an impasse resulting from the Articulação s refusal to give the Secretary General position to AE member, Arlindo Chinâglia. As initially promised, Cândido Vaccarezza received the appointment in exchange for having defected from AE prior to the X Encontro. The 1996 municipal elections did not alleviate internal tensions. After the elections, an anonymous moderate petista in the CEN raised the ire of radical petistas, arguing in agreement with a recent Estado de São Paulo article that moderates had fared better than radicals, and that the PT would benefit from left tendência s domestication and maturation. 246 Reflecting on the X Encontro and resulting impasse in forming the CEN, Hamilton Pereira recalled years later: See the point we reached in terms of fragility Lacerda (2002: 66-9) and Ribeiro (2010: 207-8) concur that, at this moment, the PT came as close as ever to a major schism. Yet internal crisis gave way, relatively quickly, to moderate hegemony. After their 1995 victory, the Articulação and DR formalized their alliance under the new name, Campo Majoritário and set out to institutionalize moderate control of the PT, which in the previous couple of years seemed to have depended on Lula s informal, behind-the-scenes machinations. 248 Campo Majoritário would remain the PT s dominant coalition into the 2000s and 246 Internal PT text, Saem os sem-voto, entra o PT adolescente, by Sonia Hipólito. From Edgar Leuenroth archive. 247 Interview with Hamilton Pereira in de Moraes and Fortes (eds., 2008: 273). 248 Despite an organizational format of low leadership autonomy, Lula found ways to exert authority over the party and to assume independent or parallel actions when he could not do so. Nonetheless, it was evident to party moderates that they could not always depend on such informal measures. After the

169 transform the PT into a center-left, electoral-professional, and (according to Ribeiro 2010) internally undemocratic party. The fundamental sources of PT unity Though beset by internal conflict from inception, the PT, unlike other Brazilian parties (e.g., PDS and PMDB) and other Latin American new left parties (e.g., IU, LCR, FMLN), never suffered a major schism. 249 What factors, fundamentally, enabled the PT to avoid debilitating schism? Petista socialism and internal democracy? Many party members and observers have pointed to the party s undefined socialism and participatory democracy. While the former made the PT s ideology contestable, the latter gave members and factions a fair chance at influencing the evolution of this ideology. On the importance of ideological indefinition for cohesion, Ribeiro (2010) writes that [t]he maintenance (and even exaltation) of this ideological indefinition (nebulosidade), over the years, became an incentive for tendências to remain in the PT. The petista ideology was seen as terrain in dispute (campo em disputa) by the factions, each with its more consolidated theory. A more precise ideological definition would be an incentive to defect for disagreeing correntes (Ribeiro 2010: 214). Along similar lines, Marxist petista Hamilton Pereira, in reference to the texts approved at the I Congress (1991), opined that socialismo petista...[was] a lucid, intelligent formulation because it [didn t] opt for a conceptual or ideological definition of socialism, but suggest[ed] a path to follow in order to get a formulation. That was a political accomplishment of the greatest importance... [The PT] managed, presidential defeat they would take steps toward developing more institutionalized means to secure their leadership position within the organization (Hunter 2010: 123). 249 Some might regard the 2004 schism, in which Marina da Silva defected from the PT to form PSOL. PSOL, however, quickly became a marginal force, and the PT won the next two presidential elections. In contrast, the offshoots of the PDS and PMDB the PFL and PSDB became institutionalized and, on most accounts, surpassed their predecessors. 161

170 in the thick (miolo) of the crisis of socialism s demoralization, to find a formula that [gave] ideological cohesion to the party. 250 Some of the same scholars (e.g., Ribeiro 2010), as well as numerous petistas, have also suggested that the PT s much vaunted internal democracy, by giving factions a fair shot at shaping the party s explicitly open-ended future, helped the PT avoid debilitating schisms. Internally, the PT constituted a forum in which a wide range of individuals and groups could dialogue, debate, and vie for influence over the party s direction. Tendências competed in elections at the zonal, municipal, state, and national levels, and the party eventually allocated executive functions, at all levels, through proportional representation. Scholars and petistas claim that this internal regime fostered a sense of inclusion and agency among participating petistas. In particular, scholars and party members have identified the implementation of full proportionality in the CEN as decisive for avoiding major radical defections during the second half of the 1980s (e.g., Ribeiro 2010). The ill-defined, open-ended nature of petista socialism did indeed allow for internal ideological contestation, and there is no question that the degree of internal democracy and level of participation in PT far outstrip[ped] that of any other important Brazilian party (Keck 1992: 121), 251 Yet the exogenous contribution of these variables (ideological indefinition, internal democracy) to PT cohesion should not be overstated. While a more specific party ideology might have alienated important sectors within the PT, petista socialism possessed far too little positive content to fuel the PT s heterogeneous base, absent other factors. Internal democracy-centered explanations of PT cohesion overlook a number of key counterarguments. First, the PT only implemented proportional representation at the national executive level in 1990, a full decade after the party formed. Until 1990, thus, PT mod- 250 Interview with Hamilton Pereira in de Moraes and Fortes (eds., 2008: 270). 251, and that the PT has initiated into political life thousands of cadre who feel confident about publicly defending party positions 162

171 erates monopolized the party s executive functions at the national level, dominat[ing] the strategic posts (Ribeiro 2010: 211). Second, although proportional representation in the national executive kept radical factions in the PT, the defection of radical factions, in all likelihood, would not have seriously weakened the party. The PT would have retained its core leadership, including Lula, and the party brand would have remained intact. Third and finally, internal democracy arguments focus on institutional design but do not account for institutional strength. Comparative evidence suggests that, with respect to internal procedures for adjudicating conflict, the particularities of institutional design matter relatively little. 252 In order for internal institutions to act as a source of cohesion, all major players must buy in. From the PT s inception, the tendências, by and large, accepted internal election results and complied with internal decisions made by opponents in power. This requires explanation. In other important parties and coalitions, factions have refused to submit to the dictates of internal opponents. In Peru s Izquierda Unida (IU), for example, every constituent party reserved veto power in the party cúpula (see Chapter Five). In Mexico s PRD, accusations of fraud continue to mar internal elections between party corrientes (see Chapter 3). The strength of the PT s democratic institutions simply raises the question: Under what conditions do ideologically heterogeneous politicians and activists internalize the ethos of opposition loyalty expressed in such petista statements as: I ve always held that the tendências are important, but they cannot be more important than the party organs. Thus, any position taken by the tendências regarding governance must pass through the party organs... [R]espect for pluralism and internal diversity is a sacred thing for us. Olívio Dutra In Mexico s PRD, most internal compromise and conflict adjudication occurred through through the office of party leader, Cuauhtémoc Cárdenas. Although the PRD possessed an internally democratic formal structure based on local, state, and federal elections and proportional representation, the party s parallel, informal hub-and-spokes regime, based on Cárdenas s active brokerage of competing demands, accounted for virtually all major decisions. Chapter 3 presents this argument in depth. 253 Interview with Olívio Dutra in de Moraes and Fortes (eds., 2008: 130-1). 163

172 We learned democratic practice...with the factions inside the PT. [The factions] helped a lot in building and consolidation the PT. Did we disagree? We disagreed. Are we going to argue? Let s argue. Are we going to resolve our differences? We re going to resolve them. Are we going to come up with a plan of action? We are. This was always respected, and it was very good for the PT. Djalma Bom 254 The great virtues of the PT organization were its internal democracy, its openness, without rigidity, non-sectarian, non-dogmatic, accepting of currents within it that came from the Church, currents of every type, that wanted to fight against the dictatorship. These are the things that make the PT unique... But we ll never cease to have a nucleus, a current that thinks more about the future, that thinks what the party doesn t think... Raul Pont 255 If not ideological indefinition or internal democracy, what factors fundamentally account for the early PT s cohesion? Lula s indispensability: electoral coattails and internal legitimacy For PT radicals, the prospect of winning national power of defeating the bourgeois state provided a strong electoral incentive to remain in the PT, and thus helped compensate for ideological differences with moderate petistas. In a June 1987 internal text, On partybuilding, Marxist petista Raul Pont wrote, If the PT s organs of debate and deliberation are democratic, and factional rights are assured, it is imperative that once [internal] decisions are made, the party acts cohesive and unified. This unity is not dictated by some desire of ours or the whim of more orthodox spirits. It is dictated by the unity and strength of our enemy. The bourgeois state, through monopolies, the police, the armed forces, and the media, is increasingly unified in its repressive actions, in its capacity to act. 256 Yet for petistas, the possibility of winning national power, during the formative phase, did not come primarily from the PT brand. It came primarily from Lula s membership in the party. Like most new parties, the PT did not possess a powerful, solid brand upon creation. 254 Interview with Djalma Bom in de Moraes and Fortes (eds., 2008: 98). 255 Interview with Raul Pont in de Moraes and Fortes (eds., 2008): From Edgar Leuenroth archive. 164

173 The party won less than three percent of the Chamber vote in 1982, roughly five percent in 1986, and a little over ten percent in Especially given the electoral clout of the ideologically proximate PMDB, the PT label provided a relatively weak electoral incentive against elite defection. To a significant degree, thus, the Lula brand substituted for a strong PT brand during the formative phase. From the beginning, early petistas coordinated, in large measure, around the goals of electing Lula governor of São Paulo (in 1982) and, subsequently, president of the Republic (in 1989 and 1994). On this point, party members and observers leave little room for doubt. Marxist petista Hamilton Pereira summarized that the PT confronted enormous difficulties to become consolidated as a political pact in the 1980s... The PT achieved an elevated padrão of discipline and unity of action. In large part generated by the expectation to elect Lula president, it s true. 257 All PT candidates, Lincoln Secco summarized in an interview, depended on Lula s electoral performance, which pulled along lower-level PT candidates. 258 Pedro Ribeiro observed in conversation that everyone knew [Lula] was the name with the greatest electoral appeal in the PT. 259 Keck (1992) observes,...[i]t was recognized that the party had to include Lula to get off the ground (Keck 1992: 81). Hunter (2010) argues that [t]he PT benefited from having a single virtually irreplaceable leader who enjoyed more societal support than the party (3). In short, because Lula enjoyed greater external popularity than the PT, his presence provided a strong electoral incentive for petistas to remain within the party fold. Crucially, however, it was the risk of internal radical takeover, not the threat of radical defection, that produced the early PT s most significant internal crises. After both the 1982 electoral debacle and, years later, the left tendências surge in the mid-1990s, Lula and his 257 Interview with Hamilton Pereira in de Moraes and Fortes (eds., 2008: 270). 258 Interview with Secco. 259 Author s conversation with Ribeiro,. 165

174 moderate allies sought, facing possible failure, to impose on the PT s radical wing. If they had failed, they might have abandoned the PT project as electorally non-viable, like other new left leaders in Latin America under similar circumstances (e.g., Alfonso Barrantes (see Chapter Five)). In describing Lula s internal dealings with the PT, many petistas and external analysts highlight the ways in which Lula stood above faction. 260 Although symbolically associated with Articulação (Secco 2011: 93) and, more generally, with PT moderates, Lula put himself above the factions in various ways and refrained from explicitly taking sides. 261 Lula chose, for example, not to join the Articulação formally. He also demonstrated prudence in abstaining, consistently, from contentious internal votes. In an academic study of the ideological tensions and ambiguities in the PT from 1980 to 1995, de Azevedo (1995) observes that Lula always avoids participating in any faction or involving himself directly with any party bloc (154). In an extension of this argument, Marxist petista Hamilton Pereira, in 2008, extolled Lula and his moderate allies for standing above and embracing the PT s heterogeneous factions, and for defining the PT very clearly, at the outset, as a political rather than ideological pact. 262 Yet these claims should not be overstated. During the internal crises mentioned above, Lula sided with the moderates, and whether he did so explicitly or implicitly, the PT base knew where Lula stood. PT cohesion thus depended, in large measure, on Lula s demonstrated ability to tame, or domesticate, radical elements of the PT base. 260 Phrase borrowed from Ansell and Fish (1999). 261 Interview with Secco. 262 The great merit of our first leaders, Lula, Zé (José) Dirceu, was that they formulated the following vision: We cannot start from the premise that we are an ideological pact. For the PT to work, we have to constitute ourselves as a political pact, that abrigue (shelters, protects) these differences. But that perception was not made explicit initially. It will only be made explicit at the V Encontro. It was a kind of instinctive thing: People, don t make this a religion; that will put us on a bad course (desanda). It can t become the Church... [The leadership] was a non-group group, a tendência against tendências (interview with Hamilton Pereira in de Moraes and Fortes (eds., 2008: 264). 166

175 Lula s internal leverage came, in part, from his aforementioned coattails. In her analysis of PT adaptation in the 1990s, Hunter (2010) demonstrates that Lula, in order to demand change within the party, used his electability as leverage (3). The presence of a popular leader with a singular ability to gain societal support, she later summarizes, served as crucial leverage for change within the [PT] (6). Lula s internal leverage, however, did not stem from electoral coattails alone. Comparative evidence suggests that electoral coattails are not sufficient for a moderate party leader to impose pragmatic measures on a radical base, especially if a separate leader within the party commands greater loyalty among the radical base. 263 Unlike other leaders of new left parties in Latin America, Lula combined electoral coattails with remarkably broad-based support within the heterogeneous PT. Perhaps more than any other single event, Lula s victory by overwhelming consensus in the 1993 contest for PT president despite left tendências control of the DN demonstrated Lula s unique level of support across the PT base. Where did this support come from? In contrast to other new left leaders in Latin America (e.g., Alfonso Barrantes), Lula, in his capacity as PT leader, began with strong ties to his party s core feeder groups, having led and actively participated in the PT s autonomous labor struggles of origin. By the end of the 1970s, Lula, to a greater degree than any other individual, had developed collaborative relationships with the full range of leaders involved in the autonomous labor movement, from rural, industrial and middle-sector unionists to Catholic leftists and Marxists. As Sluyter-Beltrão (2010) summarizes in his study of the movement s rise and decline, the metalworkers of São Bernardo, led by Lula, constituted...the principal hub of the New Unionism (Sluyter-Beltrão 2010: 3; emphasis added). Since the early PT grew out of the new unionism, Lula constituted the hub of the early PT network as well. 263 By way of illustration, Peru s IU split apart and collapsed before the 1990 presidential election because party leader Alfonso Barrantes, on whom the IU depended electorally, failed to impose upon the IU s radical elements and exited the party in order to run on a separate ticket. See Chapter Five. 167

176 More fundamentally, however, Lula s strong ties across the PT base, both before and after the PT s creation, were rooted in internal legitimacy and, for many, mística. 264 The early PT s animating narrative the founding myth that motivated early moderates and radicals cast the PT as the first authentically popular political expression in Brazil s history. The PT s São Paulo nucleus of new union leaders, led by a manual laborer who had lost a finger in a metalworking accident, 265 incarnated this myth. Lula and the São Paulo autênticos made the PT narrative credible. Ribeiro (2010) summarizes that the paulista nucleus acted...as a symbolic amalgamation, providing the collective incentives fundamental to party-building. In particular, the charismatic figure of Lula, the líder máximo do PT, was the party-building project s main source of identification and unity (251). In a conversation with the author, Ribeiro characterized Lula as the most authoritative historical voice of the party; the incarnation of PT history. 266 Unprompted, one petista observed that Lula, became a unifying force in the PT because he incarnated the myth of the party as an authentic expression of the popular sectors. 267 Along similar lines, Keck (1992) writes that while Lula was not the only leader with national credibility involved in the creation of the PT, he was the key figure. As the labor leader primarily responsible for sparking the campaigns and strikes that increased the power 264 On several accounts, Lula s personal warmth also facilitated relationship-building across moderate and radical networks. Manoel de Conceição, founding PT member, leader of the PT de Aço, and leader of the clandestine Maoist PT feeder party, Ação Popular, recalled: All my life I have had a good relationship with Lula. He has always been my friend, my brother (interview with Manoel de Conceição in de Moraes and Fortes, eds., 2008: 71). Lincoln Secco summarized that Lula is very sentimental. Despite not supporting the partisan left, he often praised (elogiava) certain people. I remember that he praised Ronald Rocha of the Tendência Marxista (which is no longer in the PT today)... In a country like Brazil, personal relationships mix easily with political relationships, even though they don t put an end to ideological differences. Some of Lula s old friends moved to the left [wing] of the party in certain phases of its history (interview with Secco). 265 In an interview, Lincoln Secco emphasized this aspect of Lula s biography in analyzing Lula s internal charisma (interview with Secco). 266 Author s conversation with Ribeiro,. 267 Interview no

177 of the whole Brazilian labor movement, his was the voice needed to give legitimacy to the formation of a party (Keck 1992, p. 77). Lula s background earned him the admiration of leading radicals (e.g., Florestam Fernandes) and moderates alike. 268 One Marxist ex-petista observed that although most Marxists considered Lula and the new unionists insufficiently radical, they admired and respected the PT leadership s authentic working-class origins. Keck (1992) observes: However much the left might criticize what it called his vacillation, it recognized that Lula was still the authentic working-class leader par excellence, and there would be no Workers Party without him (81). Lula did not impose on PT radicals by lording his external appeal, or his control of union resources, over them. The manner of Articulação s rise illustrates this point. The São Paulo new unionists who led the Articulação did not have significant resources at their disposal. Further, the 1982 electoral debacle had chastened them above all Lula, who had finished a distant fourth in the São Paulo gubernatorial race. Yet they quickly established internal dominance, formally uniting, allying with a broad range of leaders from the Catholic and Marxist lefts, and, under the Articulação label, winning repeated landslides in internal elections. At root, the creation of the Articulação, in Keck s (1992) formulation, was an attempt to promote a particular definition of the party s essence... [Articulação s] legitimacy derived primarily from its ability to shape a credible vision of the party s identity. Its centrality reinforced the argument that despite the sectarian divisions...the core or essence of the PT was composed of people who brought to the party a wide range of experiences in popular struggles (116). Without Lula, the working-class leader par excellence (Keck 1992: 81) and ultimate symbol of the PT s original struggles, the Articulação, symbolically led by Lula (Secco 2011: 93), 268 Interview with Secco. According to Secco, Lula lost much of his respect in 2002, when he promoted a conservative economic platform and enlisted the campaign services of electioneer, Duda Mendonça (interview with Secco). See also: Interview with Hamilton Pereira in de Moraes and Fortes (eds., 2008: 268). 169

178 could not have claimed the party s moral center. The Articulação s moral authority followed from the central role that Articulação s leaders, above all Lula, had played in the popular struggles that gave birth to the PT. Not only had Lula cut his teeth in popular struggles prior to the PT s creation, he remained committed to the PT s civil society linkages, both rhetorically and in practice, throughout the formative phase. To begin, Lula continued to build internal linkages after the PT s creation, regularly visiting local party organs in the early years (Secco 2011: 59) and, throughout the 1980s and 1990s, helping bring new civil society leaders into the PT. When the occasion merited it, Lula also weighed in, publicly, against the PT s more electoralist and institutionalist currents throughout early development. After a group of victorious PT candidates tempered their rhetoric in the late 1980s, for example, Lula rebuked them, warning of those who only knew the red carpet of parliament and had not slugged it out (amassar barro) in the slums. 269 In the I Congresso (1991), Lula stressed on repeated occasions that [w]e must not let electoral concerns take over the party s agenda. 270 In short, a leader without Lula s popular mystique, organic connections, and ongoing commitment to the grassroots may not have succeeded in taming the PT base, whether in the 1980s, when Lula lent decisive weight to the creation and consolidation of Articulação, or in the mid-1990s, when Lula won the party presidency and helped engineer the PT s long-term shift to center-left electoral-professionalism. * * * In summary, Lula acted as a fundamental source, perhaps the fundamental source, of unity in the early PT. Not only did he furnish electoral incentives against defection when the party 269 Interview with Secco. 270 Machado and Vannuchi (1991: 6), cited in Hunter (2010: 25). 170

179 brand remained relatively weak, he also leveraged his electability and strong crossfactional ties, the latter rooted in moral authority, to balance the PT s diverse currents and, at crucial junctures, impose moderation on a largely radical base. Numerous petistas and PT analysts converge in this assessment. When asked why the early PT did not split apart despite internal divisions, Lincoln Secco, stated, unprompted, that the early PT was lucky to possess, in Lula, a leader who combined external and internal charisma. 271 In an interview with the author, Frei Betto, when asked what united the PT, stated that Lula was always the unifying factor. Betto stressed that even when certain tendências disagreed with Lula, they generally did not speak out publicly against him. 272 When asked the same question, early petista and current São Paulo vereador, Antônio Donato, stated that the early PT was the most viable electoral vehicle on the left, but also that Lula succeeded in creating a strong center without squashing (esmagar) the minorities. 273 Pedro Ribeiro, in an conversation with the author, remarked that Lula was always one of the few glues between the factions, above them all, unifying the party... Lula was a great negotiator of agreements, and also a guarantor of those agreements - and not only due to the electoral factor, but because he was the most authoritative historical voice of the party, the incarnation of PT history. 274 Rodríguez (n.d.) concurs in her comparative study of the PT and Mexico s PRD, summarizing that Lula, like Cuauhtémoc Cárdenas, was a strong [leader]...who centralized the functions of the party voice [and], due to [his] moral quality, established [himself] as [a guarantor] of pacts between groups (208). In her comprehensive study of PT adaptation in the 1990s, Hunter (2010) attributes Lula s 271 Interview with Secco. 272 Interview with Frei Betto. 273 Even when the DN expelled the Trotskyist Convergência Socialista and Causa Operária in 1992, Donato observed, they did so against their own will. By violating the regulamentação de tendência, Donato argued, CS and CO, in effect, chose to leave the party (interview with Donato). 274 Author s exchange with Ribeiro,. 171

180 successful imposition on radicals to his combination of external and internal support: [A]s a founding member and longstanding leader within the PT, Lula had the authority to convince a wide range of party militants that he could win presidential office if only they would go along with the measures he advocated for change... His candidacy as party president met with overwhelming internal consensus despite fierce conflicts over other issues. At the end of the day, it was Lula s popularity among petistas [emphasis not in the original] and the widespread belief that he was the candidate the PT could put forth that brought the party back into his fold (36, 122). Hunter (2010) concludes, more generally, that the existence of a single leader who enjoyed both a strong presence within the party and popularity with the electorate distinguishes the successful PT from other failed left parties and coalitions in contemporary Latin America (3; emphases not in the original). 275 The current chapter, in conjunction with Chapter Five, elaborates a similar argument. Chapter Seven will argue that Peru s new left coalition, Izquierda Unida (IU), collapsed by schism due, fundamentally, to the internal weakness of moderate coalition leader Alfonso Barrantes. Barrantes situation in the IU prior to the 1990 Pervuian presidential election roughly paralleled Lula s situation in the PT prior to the 1994 Brazilian presidential election. Among IU elites, Barrantes alone stood a chance of winning the presidency, but he did not possess moral authority or organic ties within the coalition. IU radicals had always wavered in their support for him, and in the lead-up to the presidential elections, it appeared likely that Barrantes would lose in a closed primary for the coalition s presidential nomination. Barrantes demanded the nomination outright, absent a primary, and IU radicals refused. Instead of submitting to radical demands and running in the primary, Barrantes split from the IU to run for president in 1990 on a separate ticket. The IU collapsed shortly thereafter In this connection, Hunter (2010) identifies Felipe González as an interesting parallel of someone who drew on his popularity among the electorate and within the party to gain acceptance for measures that encountered strong internal opposition (36). 276 Hunter (2010) concurs with this assessment in a brief conclusion: Barrantes proved unable to persuade elements outside his own wing to follow his lead. Unlike Lula, he could not manage to leverage his 172

181 popularity into moderating the IU s platform and image to sustain it as a serious electoral contestant. Whereas Lula was a founding member of the PT and remained at the heart of the party, Barrantes did not enjoy this standing. His independence may have allowed him to stay above the fray, but it also hindered his efforts to exert influence over parties within the Izquierda Unida. The parties of the far left never fully accepted Barrantes leadership. Javier Diez Canseco of the PUM, a key component of the IU s radical wing, even charged that Barrantes had reduced the IU to an electoral front around a caudillo (194). 173

182 Chapters 4-5 The Survival of Mexico s Party of the Democratic Revolution Chapters Four and Five analyze the survival of Mexico s Party of the Democratic Revolution (PRD). Founded in 1989, the PRD has become Mexico s leading left party and a perennial contender for national and subnational power. Since the late 1990s, the PRD has won eleven gubernatorial elections in six Mexican states and continuously held the jefatura de gobierno of the Federal District, the second most important elected office in the country. The PRD has elected at least ten percent of Mexico s federal deputies in every congressional election since 1994 (Figure 1). In 2006, the PRD elected a quarter of Mexico s federal deputies, and the party s Andrés Manuel López Obrador came very close to winning the presidency. Given Mexico s size, economic strength, and geopolitical significance, the PRD is one of the most important new left parties in Latin America. Figure 4.1: The PRD in the Chamber of Deputies (1991-present) 174

183 Despite the PRD s enduring, national electoral significance, leading scholars have often treated the party as a failure. Greene (2007) emphasizes that left ideologues in the PRD leadership and base have made electorally suboptimal choices and thus prevented the party from winning the presidency. 1 Bruhn (1998) argues that the early PRD did not consolidate, failing to coordinate its activists, build stable party/interest group relations, and construct a strong ideological identity (198). 2 Hilgers (2008) similarly claims that the PRD, by engaging heavily in clientelism, has failed to develop a strong programmatic identity (Hilgers 2008). Rodríguez (n.d.) characterizes the PRD as internally dysfunctional, emphasizing that the PRD, unlike Brazil s Workers Party (PT), has weakly institutionalized factional relations and, in her view, no stable dominant coalition (548). 3 These chapters do not contest the above accounts or critical accounts provided elsewhere. 4 Yet they do begin from a different and insufficiently explored premise: that the PRD, if viewed in broader comparative perspective, is a clear case of new party survival. Despite electoral setbacks and a certain degree of internal dysfunction, the PRD is one of a handful of new left parties in Latin America to become institutionalized and regularly contend for national power. Why did the PRD succeed in becoming institutionalized as a major national party? 1 One such choice is the nomination of Cuauhtémoc Cárdenas as the party s presidential candidate in Bruhn argues that Brazil s Workers Party (PT), in contrast, did consolidate. 3 On the latter point, Borjas (2003) disagrees. 4 Scholars have argued that the PRD has governed poorly at the subnational level (Borjas 2003; Rodríguez n.d.). Ortega (2010) implicitly treats the PRD as an organizational failure, highlighting the party s infrastructural weakness in various Mexican states and at the base level more generally (13, 17). Martínez (2005) highlights various forms of dysfunction within the national PRD organization. For additional critiques, see Sánchez (2008, 1999). 175

184 Chapter 4 Building the PRD: adversity and survival Castañeda (1995) has compared Mexico s 1994 presidential election to a soccer game where the goalposts were of different heights and breadths and where one team included eleven players plus the umpire and the other a mere six or seven players (131). One could apply this metaphor to the PRD s entire formative period. The PRD grew out of cardenismo, a heterogeneous mass movement to elect Cuauhtémoc Cárdenas president of Mexico in Cárdenas s bid for the Mexican presidency constituted an attempt, after nearly a decade of economic and humanitarian crisis, to defeat the authoritarian, hegemonic PRI, reverse Mexico s neoliberal turn, and restore the authentic revolutionary legacy of social justice and economic nationalism championed by Cuauhtémoc Cárdenas s father, ex-president General Lázaro Cárdenas. After a presidential campaign marked by extreme resource asymmetries and systematic mass media bias against the left, Carlos Salinas of the ruling PRI defeated Cárdenas in an election widely viewed as fraudulent. In the election s aftermath, the cardenista movement grew in size and intensity, and from it, the PRD emerged. During the late 1980s and most of the 1990s, the PRI enjoyed a vast public and private resource advantage over its competitors, especially the PRD. Until the mid-1990s, the PRI monopolized broadcast media and controlled the instruments of fraud and repression. The PRI used all of these resources and tools to keep the PRD, in particular, in the political wilderness. The PRD thus developed without patronage or financial resources, and in a closed, hostile media context. The party could not achieve electoral progress except through extensive volunteer activism. During the initial years of party development, national PRD leaders, acutely aware of their disadvantages, made it a top priority to develop party activist networks in as much of the national territory as possible. 176

185 Lacking the resources and institutional capacity to finance or administer large-scale territorial implantation from the top down, the national PRD organization depended on the bottom-up initiative and dispersed human and infrastructural resources of the local leaders, movements, and organizations sympathetic to cardenismo. Just a couple of years into the PRD s existence, thousands of local party nuclei had sprouted up, in various regions of Mexico but especially the Federal District and poor southern states. Most local party organizers had origins in extraparliamentary social left movements and organizations, defecting PRI networks, and the traditional Marxist political left. To build local party branches, these local organizers recruited activists and used physical locales and administrative resources from their feeder structures of origin. The PRD s bottom-up, decentralized organizational expansion quickly resulted in a high degree territorial implantation. By the middle of the 1990s, the PRD claimed over one-million members and formal offices in over half of Mexico s 2000-plus municipalities. The PRD activists who made up the core of the party organization possessed high levels of commitment. Early adversity selected for these committed activists. Due to mass media hostility, fraud, and acute penury and resource asymmetry, party elites could not offer salaries to activists, or make credible promises of national electoral success and patronage spoils in the short term. Moreover, repression and murder at the hands of subnational PRI authorities raised the cost of activism in the PRD s southern strongholds. Between 1988 and the mid- 1990s, hundreds of FDN and PRD activists lost their lives, and thousands spent time in jail. Given these conditions, pragmatists and careerists were unlikely to join the PRD. Only believers, or individuals with sincere ideological commitments to the party, would participate in such a poorly resourced, electorally uncertain, and personally risky political undertaking. 5 5 Term from Panebianco (1988: 26-30). 177

186 Electoral disadvantage and repression selected for but did not generate committed PRD activists. The polarization of the late 1980s produced the higher causes that fueled the PRD s early volunteer army. The goal of taking the reins of national government from the PRI, and dealing a decisive blow to authoritarianism and neoliberalism in the process, spurred local leaders and organizations from across the left ideological spectrum to join in building and campaigning for Cárdenas and the early PRD. Activist commitment solidified when the PRI, in order to retain power, violently repressed and systematically defrauded the left at the national and subnational levels. Early perredistas derived moral and psychological benefits from their active contribution to the party: benefits that compensated for unremunerated time and labor and, in some cases, the threat of violence and even murder. In sum, within a few years of its creation, the PRD possessed a strong and durable organization, with official party branches in large swathes of the national territory, and with activists driven by a belief in the PRD s cause, not by a short-term material or electoral payoff. Organizational reach and activist commitment fortified the early PRD amid successive electoral crises. After both the disastrous 1991 midterms and Cárdenas s unexpectedly lopsided loss in the 1994 presidential election, party activist networks rebounded in territorial bastions. Setting the stage: PRI decline and the creation of the PRD Public support for Mexico s dominant PRI fell precipitously in the 1980s. Latin America s 1980s debt crisis devastated Mexico, ushering in nearly a decade of economic crisis. Mexico s per capita income decreased for four of the years between 1982 and 1987, including by nearly six percent in Inflation increased at an average rate of 88.4 percent per year between 1982 and 1988, reaching 159 percent in The real minimum wage fell by roughly half between 1983 and The ranks of unemployed swelled over the course of the decade, as 178

187 public-sector employees lost their jobs and small businesses folded. 6 Like most Latin American governments, the administration of Miguel de la Madrid (PRI) reacted to the 1980s debt crisis by shifting from economic statism and protectionism to austerity and free market reforms. In the initial years of de la Madrid s sexenio ( ), the PRI government implemented a series of fiscal and market reforms including tariff reductions, large privatizations, broad wage freezes, and draconian budget cuts. The shortterm social costs of recession, inflation, and fiscal tightening proved immense. As the decade closed, middle- and lower-class Mexicans had fewer jobs, lower incomes, and fewer public benefits to compensate for unemployment or reduced income. Separate, simultaneous developments caused further reductions in PRI support. The PRI s core rural constituencies continued to shrink with urbanization. Highly publicized corruption scandals fed popular discontent, particularly given the economic circumstances of the 1980s. Crucially, in 1985, the Federal District suffered a devastating earthquake, which destroyed parts of the city and killed over 10,000 individuals, predominantly from the lower class. 7 The de la Madrid administration s earthquake response was widely perceived as incompetent and inadequate. The combination of unfavorable demographic shifts, prolonged economic crisis, corruption scandals, and humanitarian catastrophe generated the PRI s most significant crisis of legitimacy since the late 1960s. Between 1983 and 1987, public support for the PRI fell twenty-five percent, largely due to increased disaffection among Mexico s popular sectors (Bruhn 1998: 119). This shift in public opinion created an electoral opportunity for the PRI s challengers, particularly those on the left. 6 The figures in this paragraph are taken from Bruhn 1998: 118 and Ortega 2008: One PRD founder recalls that the most significant drop in PRI public support among Federal District voters came as a result of the earthquake, creating a space for a left electoral alternative (interview with Saucedo). 179

188 In 1986, a group of dissident PRI leaders founded an anti-neoliberal faction, the Corriente Democrática (CD). Led by Michoacán governor Cuauhtémoc Cárdenas, former PRI president and UN ambassador Porfirio Muñoz Ledo, and former UN ambassador Ifigenia Martínez, the CD proposed to rebuild the PRI s eroding societal linkages and reverse de la Madrid s austerity and privatization through debt repudiation, wage hikes, increased protection of domestic industry, and increased state involvement in production. CD elites sought to advance their cause from within by democratizing the PRI. They pressed for more vigorous, wide-ranging programmatic debate and the implementation of formally democratic candidate selection procedures (Rodríguez n.d.: 185-6). The PRI did not carry through substantive internal democratic reforms, and Cárdenas did not receive consideration as the PRI s presidential candidate. In October of 1987, Carlos Salinas de Gortari, de la Madrid s Secretary of Program and Budget and an enthusiastic proponent of de la Madrid s neoliberal adjustments, was fingerpicked as the PRI s presidential candidate (Rodríguez n.d.: 186). The CD defected from the PRI in protest of the party s internal authoritarianism and, programmatically, its technocratic neoliberalism and abandonment of revolutionary goals and ideals. Less than two weeks later, Cárdenas announced his candidacy on the ticket of the Authentic Party of the Mexican Revolution (PARM), formerly a PRI satellite party, or parastatal. 8 Soon after, two additional parastatals, the Socialist Popular Party and the Socialist Workers Party, 9 broke with the PRI and joined the CD and PARM to form the National Democratic Front (FDN). Several parties and organizations promptly joined the FDN, including the 8 The parastatals were a group of electorally marginal, nominal opposition parties which, prior to the PRD s formation, had cooperated with the PRI to the point of co-nominating PRI presidential candidates (Bruhn 1998: 3). Months before officially backing Cárdenas, the PARM announced that it would no longer support PRI presidential candidates (Rodríguez n.d.: 186). 9 PPS, PST. After deciding to back Cárdenas, the Socialist Workers Party (PST) changed its name to the PFCRN. 180

189 Social Democratic Party, the Revolutionary Socialist Party, the Green Party, Democratic Unity, the Progressive Forces of Mexico, and the National Workers and Peasants Council. In January of 1988, the FDN officially nominated Cárdenas as its presidential candidate. Cárdenas possessed a powerful electoral brand, particularly among peasants and urban popular sectors. This brand derived from his unique lineage. As the only son of General Lázaro Cárdenas, Cárdenas symbolically bore, for many Mexicans, the authentic legacy of the Mexican Revolution. General Lázaro Cárdenas had served as a general in the Mexican Revolution and as president of Mexico from 1934 to As president, he had left a lasting imprint, implementing extensive agrarian reform, making pioneering investments in education, infrastructure, and social insurance, and nationalizing Mexico s oil industry. PRI presidents before Salinas uniformly recognized the General as Mexico s greatest ever (Borjas 2003: 510). Early campaign events drew an explicit connection between Cárdenas the son and Cárdenas the father. In the initial months of 1988, the Cárdenas campaign held a series of landmark mobilizations. One took place in the Comarca Lagunera of Coahuila and Durango, where masses of peasants had benefited from Lázaro Cárdenas s land reform policies. Another took place in Mexico City s zócalo, where Cárdenas led a rally to celebrate the fiftieth anniversary of his father s nationalization of Mexico s oil industry. 10 While Cárdenas s last name evoked his father and the Mexican Revolution, his first name, Cuauhtémoc, evoked Mexico s pre-conquest Aztec history. Cárdenas s physical traits and grave, austere attitude further reinforced the indigenous prototype rather than the mestizo or creole one (Borjas 2003: 293). Cárdenas thus concentrated many of the symbols of national identity displaced during [Mexico s] recent history (Borjas 2003: 293). Bruhn 10 Another took place at Mexico s UNAM, a large university and bastion of Mexico s Marxist left (Bruhn 1998: 131-2). 181

190 (1998) observes that no other actual or potential opposition candidate could have drawn upon the legitimacy of these symbols in quite the same way (127). As of the late 1980s, the left opposition could point to the deepening immiseration of Mexico s urban and rural poor and charge the PRI with a kind of betrayal the acceptance of an externally imposed economic model at the expense of social justice for Mexicans. As governor of Michoacán (1980-6), Cárdenas had opposed, and to some extent defied, the PRI s neoliberal shift during the 1980s. Thus, the left opposition could propose, through Cuauhtémoc Cárdenas, a return to the revolutionary nationalism practiced in the candidate s political life and embodied in his very name. 11 Cárdenas s candidacy generated mass support in various regions of the country. This support came primarily from three broad groups: the cardenista left, the traditional Marxist left, and the social left. 12 The cardenista or independent left, the largest of the three groups, included citizens who supported Cárdenas but did not participate actively in separate left parties or organizations. Like Cárdenas, many were PRI defectors and did belong to local networks of fellow ex-priístas. The traditional Marxist left, alternatively the historical or parliamentary left, encompassed a heterogeneous array of Marxist parties that, in the 1970s and 1980s, had moderated, committed to institutional politics, and sought a gradual transition to democracy through electoral participation and political reform. From the mid-1970s to the late 1980s, parties of the traditional Marxist left had created a succession of electoral alliances that culminated in the 1988 founding of the Mexican Socialist Party (PMS). Initially, for the first half of 1988, the PMS fielded its own presidential candidate, Heberto Castillo of the constituent Mexican Workers Party (PMT). The PMS refused to join the 11 Summarizing the appeal of cardenismo, one scholar observes that the FDN, unlike the opposition PAN, represented the official ideology of social justice from a group of renegade elites capable of revitalizing revolutionary ideals at a moment when...lower sectors felt abandoned by the state Brachet-Márquez (1996: 121, emphasis added). 12 Hernández (2010: 38) identifies these three groups as the three main vertientes of the PRD. 182

191 FDN on ideological grounds; unlike much of the FDN, the PMS retained the basic ideals of socialism and only begrudgingly accepted the tenets of democracy (Greene 2007: 93). Ultimately, however, practical electoral considerations trumped ideological differences. In late May of 1988, the powerful student organization, the University Student Council (CEU), expressed overwhelming support for Cárdenas at a rally in UNAM s Ciudad Universitaria. Fearing political marginalization, 13 recognizing Cárdenas s unique potential to put the left in the presidency, and somewhat amenable to Cárdenas s revolutionary nationalism, 14 the PMS, most of whose members lacked pre-pms ties to Heberto Castillo, opted to back the FDN in early June of Castillo thus retired his candidacy in favor of Cárdenas, and the PMS joined the Cárdenas campaign. Cárdenas s third base of support, the social or extraparliamentary left, included social movements and organizations that had not previously participated in institutional politics. The bulk of these movements and organizations possessed militant Marxist origins. Following the PRI s violent crackdown on the left in the late 1960s, many Marxist leaders and activists had not moderated, instead rejecting electoral participation and committing to (sometimes violent) social struggle. A large number of social left groups originated as splinter groups from the moderating Marxist establishment. Other social left organizations, usually less militant, emerged subsequently and focused on more post-materialist issues. The social left thus encompassed a wide array of actors, including ex-guerrilla nuclei, rural unions, popular urban movements, teachers unions, radical blue-collar unions, student associations, NGOs, environmental groups, homosexual rights groups, and more. Despite facing a slew of disadvantages, the FDN mounted the most significant challenge to 13 Traditionally, Mexican socialists and communists had drawn much of their support from university students, particularly UNAM students. 14 See Martínez (2005): Borajs (2003):

192 PRI hegemony since the mass mobilizations of the late 1960s. In the lead-up to the 1988 election, the PRI-controlled Televisa and other mass media outlets defamed and ignored Cárdenas, while subnational PRI authorities employed coercion and occasionally committed murder in order to minimize FDN activism, especially in the FDN strongholds of Michoacán and Guerrero. In order to maintain power, the PRI ultimately had to resort to fraud. Cárdenas lost by a wide margin in the official tally, with a reported thirty-one percent of valid votes to Salinas s fifty percent. The FDN elected nearly twenty eight percent of the country s federal deputies and four senators, including Porfirio Muñoz Ledo and Ifigenia Martínez in the Federal District. Immediately, the FDN presented evidence of systematic voting irregularities that had favored the PRI, focusing on the presidential election. 16 Cárdenas rejected the official election results and called, unsuccessfully, for a new, clean election. By refusing to recognize Salinas as the legitimate president of Mexico, Cárdenas strengthened his image as an external, anti-system, opposition force. The Salinas administration assumed power under a cloud of illegitimacy and moved quickly to legitimate and consolidate its rule (Ortega 2008: 206). Soon after the inauguration, the PRI forged a pact with the conservative PAN, agreeing to recognize all of the party s future subnational victories in exchange for PAN legislators support of PRI economic and social policies. In an attempt to pacify the left opposition, the Salinas administration offered the regency of the Federal District to Cárdenas (Borjas 2003: 245). Although some key FDN figures (e.g., Porfirio Muñoz Ledo) favored such a pact with the Salinas administration, Cárdenas, supported by the bulk of the FDN base, rejected Salinas s offer and imposed a line of intransigencia: steadfast opposition, sustained confrontation, and refusal to negotiate with the PRI under any circumstances. 16 Official rates of electoral participation in rural areas tended to exceed participation in urban areas. In many municipalities, the total number of votes recorded exceeded the number of registered voters. Results from nearly 25,000 voting booths went unreported (Ortega 2008: 201-2). 184

193 In an already polarized country, the PRI s defamatory, violent, and ultimately fraudulent response to the rise of cardenismo further intensified social polarization and opposition mobilization. In the lead-up to the 1988 election, millions of Mexicans had taken to the streets for Cárdenas, and for the democratic and economic transformation that they believed a Cárdenas victory would signify. In response, the PRI, at the federal and subnational levels, had shown that it would stop at almost nothing to keep Mexico s left opposition from governing the country. Subnational PRI authorities had murdered left activists, and on the standard view among FDN supporters, the national PRI had encouraged, assisted in, and covered up systematic electoral fraud in order to secure Salinas s victory. Following the 1988 elections, in which fraud extended to the legislative elections as well, Cárdenas supporters dialed up their use of mass mobilization, carrying out mass marches, rallies, and public sitins, disrupting long segments of highway, and occupying city halls and government offices in various parts of Mexico. Protests were strongest in Michoacán, where the FDN enjoyed the greatest mobilizing capacity (Borjas 2003: 340). The Salinas administration and mass media characterized these actions as intentionally destabilizing (Borjas 2003: 339). In order to create an institutional channel for this energetic mass movement, Cárdenas and the top FDN leadership, in late October of 1988, proposed the creation of a new left political party. Cárdenas appealed to the millions of Mexicans who had supported his presidential bid, arguing that hastily constructed, poorly organized electoral fronts like the FDN could not effectively wage the post-electoral struggle or, more broadly, generate progressive change in Mexico. 17 Cárdenas s strong performance in the 1988 election, despite manifold disadvantages, suggested that in the next presidential election, he could win. On the Mexican left, the prospect of a Cárdenas victory and, with it, a complete transition to democracy and the beginning of a neoliberal reversal animated individuals and groups of all stripes. The 17 Martínez (2005: 58); Cárdenas (1988: 5-9). 185

194 CD and PMS immediately joined. The other constituent parties of the FDN - the parastatals - did not. 18 In addition to the CD and PMS, millions of independent cardenistas and more than twenty civil society organizations and social movements pledged their support. 19 The Party of the Democratic Revolution was officially created in May of To satisfy federal requirements for state assemblies and membership, the PRD adopted the registry of the PMS. 20 In the broadest terms, the early PRD constituted a left front forged in opposition to an authoritarian, neoliberal ruling party. The party was neither revolutionary nor reformist, but revolutionary nationalist and cardenista. 21 The party s leader commanded a large external voting base. Internally, the party brought together a cadre of dissident PRI elites (the CD), subnational ex-pri networks linked to these elites, the bulk of the traditional Marxist left, and the bulk of the social left. These heterogeneous actors united around a vision of democratizing Mexico and reversing the country s neoliberal turn. More concretely, they sought to win positions of national and subnational power and, in 1994, to deliver Cuauhtémoc Cárdenas to the Palacio Nacional. Low access to media and the state during the PRD s formative phase At the time of the PRD s founding, the PRI controlled the federal government, the governments of the Federal District and all the Mexican states, and the vast majority of Mexico s municipal governments. The PRD came into existence with no governors and only a handful 18 In justifying their decision not to join, the PARM cited the proposed party s links to communism, while the PPS and PFCRN cited the proposed party s weak commitment to socialism. In fact, the parastatals acted in their own short-term political interest, intending to minimize PRI hostility and prevent a loss of seats, or even a loss of party registry, in future elections (Martínez 2005: 58, note 26; Rodríguez n.d.: 248, note 37). 19 Martínez (2005: 58); Rodríguez (n.d.: 244). 20 In one party leader s phrase, the PRD was therefore just the continuation of the PMS by another name (González 2010: 259). Details on the PMS s dissolution come later in the chapter. 21 Interview with Flores. 186

195 of mayors in small, poor municipalities. In response to the PRD s policy of intransigencia (i.e., no negotiation), the Salinas administration adopted a strategy of selective democratization: while opening positions of power to the PAN, national PRI leaders and subnational PRI authorities sought to neutralize the PRD with a range of tactics. These tactics included systematic fraud, repression and murder, lavish campaigns, massive social transfers designed to erode left support, and crucially, the manipulation of mass media coverage. Together, these measures sharply curtailed the PRD s early electoral success and, consequently, access to state resources. Not until the late 1990s would the PRD begin to gain significant access to media, financial resources, and state power. Fraud against the PRD (late 1980s to mid-1990s) During the late 1980s and early 1990s, the PRD suffered repeated fraud at the municipal and state levels. In 1989 and 1990 alone, the PRI stole hundreds of mayoral elections in Mexico state, the southern states of Guerrero, Michoacán, Oaxaca, Puebla, and Veracruz, and additional states such as Baja California, Campeche, Chihuahua, Coahuila, Durango, Hidalgo, and Yucatán. 22 Municipal fraud occurred selectively. The PRI stole elections in larger municipalities but generally accepted PRD victories in smaller municipalities with paltry budgets. Of the PRD s first 111 mayoral victories, 78 took place in municipalities with fewer than 20,000 inhabitants (Borjas 2003: 363). During the first half of the 1990s, the most significant instances of anti-prd fraud occurred at the state level in PRD strongholds. In San Luis Potosí in 1991, Michoacán in 1992, Guerrero in 1993, and Tabasco in 1994, the PRD lost gubernatorial elections under questionable circumstances. In each case, the PRD, often presenting copious evidence, accused the PRI 22 Gómez Tagle (1994a, 1994b), Crespo (1995), Eisenstadt 2004, Rodríguez (n.d.), Borjas (2003). 187

196 of electoral fraud in addition to voter coercion, illegal levels of spending, and rampant votebuying. Each of these elections sparked mass protests. There is evidence that the early PRD suffered limited fraud at the federal level as well. After the 1991 congressional election, in which the PRD fared poorly, party leaders presented evidence of ballot irregularities, rural voter overregistration, and urban voter underregistration (Borjas 2003: 402-3). 23 The PRD demanded postponement of the voting, but to no avail. The PRD s unsuccessful general election in 1994 provoked similar allegations from the party leadership. In fact, however, the 1991 and 1994 federal elections, while imperfect, were significantly cleaner than the 1988 presidential election (Bruhn 1998: 254). Newspapers El País, The Financial Times, The Guardian, The Times, and Barcelona-based La Vanguardia all reported, for example, that the PRI did not commit large-scale fraud in the 1994 presidential election. 24 At the federal level, the PRD s electoral disappointments and setbacks during early development resulted primarily from factors other than fraud, including repression, stark resource asymmetries, and mass media ostracism and hostility. Repression of the PRD (late 1980s to mid-1990s) From the late 1980s to the mid-1990s, subnational PRI authorities violently repressed FDN and PRD activists in select strongholds. 25 In the 1988 election s lead-up and aftermath, thousands of FDN activists were accosted, persecuted, beaten, or jailed, and nearly fifty were killed. After the PRD s creation, subnational PRI authorities murdered hundreds of 23 See also Rodríguez (n.d.): Cited in Borjas (2003): In fact, Mexico s PRI regime and its allies committed more political violence against the left, in proportional terms, than the bureaucratic-authoritarian regimes of Brazil (Rodríguez n.d.: 297) and Uruguay. 188

197 PRD activists and organized peasants, both to quell specific instances of mass mobilization and to discourage opposition activism more generally. Sources indicate that, by the late 1990s, opponents of the left had murdered between 250 and 600 FDN and PRD activists. 26 Roughly sixty percent of these murders occurred between 1988 and 1994, the remainder between 1994 and The murders were uncoordinated and relatively dispersed, although the majority took place in poor southern states, where Cárdenas and the social left claimed large, active, fiercely loyal, and often radical bases. In these poor regions, homicides rarely resulted in arrest. The largest number of murders occurred in Michoacán, Guerrero, and Oaxaca, respectively. 28 In some instances, the PRI targeted key individuals for killing. Four days before the 1988 presidential election, for example, assailants in the Federal District murdered two of Cárdenas s top campaign aides, Francisco Xavier Ovando and Román Gil, and left their bodies to be found in an apparent warning to the FDN (Bruhn 1998: 139). In PRD strongholds, PRI groups also inflicted softer forms of repression, issuing threats, beating protesters, and making arrests on trumped-up charges (Greene 2007: 95). Political arrests of left activists numbered in the thousands, perhaps the tens of thousands, between late 1980s and mid-1990s. Because repression was widespread, FDN and early PRD activists often knew victims and, in many cases, carried out party work at considerable personal risk. 29 The PRD s resource disadvantage (late 1980s to mid-1990s) 26 González et al., eds. (2010: 66); interview with Zambrano in González et al., eds. (2010: 284); Ortega (2008: 208); Bruhn (1998: 202); Borjas (2003: 45, 341, 436); Rodríguez (n.d.: 284); Eisenstadt (1999). 27 Borjas (2003, vol. II): Murders in these three states comprised sixty-five percent of the total (Bruhn 1998: 202). 29 One PRD founder opines that the current generation of young perredistas lack an appreciation of the PRD s harsh origins: When you tell them how hard it was to build the PRD, facing the PRI government, they don t understand you because they didn t live through that period (Duarte 2010: 249). 189

198 To a greater extent than repression or fraud, however, an immense resource disadvantage hindered the PRD s early electoral progress (Greene 2007). The PRD formed outside the state and with very few connections to the economic elite or well-financed organizations. From the beginning, PRD finances fundamentally depended on the public party fund, a small pittance until the landmark electoral reforms of 1996 (Borjas 2003: 297). As late as 1994, five years after the PRD s founding, the national PRD organization could only afford fifty permanent staff (Bruhn 1998: 189). The party chronically suffered from financial shortages and a lack of resources to invest in electoral campaigns, infrastructure, and professional cuadros (Borjas 2003: 297). Well into the 1990s, PRD candidates had to finance and run their own campaigns, with little or no financial or material support from the party organization. 30 In contrast, the PRI, during the PRD s initial years, enjoyed virtually unlimited access to government funds, 31 siphoned off billions of Mexican pesos from the public treasury, 32 and received billions more from business supporters. 33 Throughout the PRD s formative years, the PRI spent huge sums on electoral campaigns. National and subnational candidates used party funds and private donations to finance traditional campaign events such as rallies and concerts, and also to purchase and distribute clientelistic goods and services, such as free meals and haircuts, on a large scale. 34 Bruhn (1998) finds that in the 1991 congressional elections, for example, the PRI, unlike the PRD, ran intelligent, ruthless, and lavish campaigns, led by a professional, well-organized, and well-financed staff (279-80). 30 In the 1989 state and local elections, for example, the PRD s National Executive Committee wrote that being a PRD candidate can be a true punishment, given that each one must obtain resources, build a team and organize activities without support or advice of any kind (Borjas 2003: 346). 31 Cornelius (1996): Oppenheimer (1996); Cornelius (2004). 33 Oppenheimer (1996); Philip (1999); De Swaan, Martorelli, and Molinar Horcasitas (1998). 34 See, for example, Bruhn (1998):

199 Throughout the first half of the 1990s, the PRI systematically outspent its competitors, especially the PRD, by massive margins. In Michoacán s 1992 gubernatorial election, for example, the PRI outspent the PRD by a factor of fifty. In 1993, twenty-nine of Mexico s most prominent businessmen donated 75,000,000 pesos or more to the PRI s coming presidential campaign; added together, these donations exceeded the budget provided to all political parties by the Federal Electoral Institute (IFE) (Borjas 2003: 426-7). 35 For the 1994 general election, the PRI spent eleven times more than the PRD and PAN combined. 36 The PRI s resource advantage went well beyond official campaign spending. Shortly after taking office, Salinas and his ministers implemented the National Solidarity Program, or PRONASOL. PRONASOL allocated federal funds to municipalities for public works, food aid, agricultural credit, land title legalization, and the construction and renovation of schools, hospitals, and transportation infrastructure. A social program of then unprecedented scope in Mexico, PRONASOL, by 1992, accounted for approximately sixty percent of Mexico s infrastructural spending and approximately half of the country s social development spending (Bruhn 1998: 219). PRONASOL was a central component of the PRI s strategy to win back public support after the cardenista surge and electoral crisis of The PRI distributed PRONASOL s vast resources in a timed, targeted, and clientelistic manner, 38 with the specific goal of un- 35 All the businessmen in attendance agreed to donate the requested sum or more (Borjas 2003: 426-7). 36 Personal communication with Ortega, June 9, 2011). 37 Molinar and Weldon (1994); Magaloni (2006): PRONASOL resources moved through an extensive patron-client network. Salinas directly appointed state PRONASOL commissioners, who in turn appointed local committees charged with distributing PRONA- SOL funds for local projects. Local PRI brokers individualized the transfers at the ground level, conditioning the distribution of PRONASOL food baskets, construction materials, and scholarships on electoral support for the PRI (Borjas 2003: 487). The PRI under Salinas also spent millions of dollars on PRONA- SOL advertisements emphasizing the PRI s association with the program. PRONASOL packages came painted with Solidarity colors the same as those of the PRI and the Mexican flag (Bruhn 1998: 234). 191

200 dermining the PRD. 39 The PRI systematically increased PRONASOL disbursements before elections, especially the 1991 and 1994 federal contests (Magaloni 2006: 149). During the 1994 general election season, PRONASOL transfers amounted to ninety-five percent of the resources spent by all parties. 40 The PRI did not disburse PRONASOL funds to PRI or PAN strongholds, so as not to waste electorally valuable resources on secure priístas or on panistas considered unpersuadable (Magaloni 2006: 141). Instead, the PRI leadership sought, through large-scale but surgical social spending, to win the votes of poor Mexicans who had supported Cárdenas in Thus, the Salinas administration concentrated PRONASOL outlays in the FDN s 1988 municipal strongholds. In order to undermine opposition party-building, the PRI made an exception in the FDN bastions where voters, after 1988, had elected a mayor from the PRD or PAN (141, 149). In these municipalities, the PRI, instead of attempting to buy back 1988 Cárdenas supporters, 41 punished them by withholding PRONASOL transfers. 42 PRONASOL bore electoral fruit for the PRI, at great expense to the PRD. The program s generous transfers held critical importance for millions of poor individuals and a multitude of underresourced civil society organizations. 43 By 1991, the PRI, through PRONASOL, had managed to co-opt or divide many of the municipal constituencies and civil society organizations previously sympathetic to Cárdenas. Hundreds of municipalities that had supported Cárdenas in 1988 defected to the PRI in 1991 and Major popular associations that 39 Molinar and Weldon (1994); Magaloni (2006). 40 Porfirio Muñoz Ledo, Proceso, no. 931, September 5, 1994, pp See Molinar and Weldon (1994). 42 Magaloni (2006): 141 and Chapter 4, passim. 43 Bruhn (1998) writes, for example, that [f]or many of the demands made by popular organizations, PRONA- SOL was the only game in town (219). 44 One party activist recalled, for example, that the Salinas administration, through PRONASOL, provided key social services such as road repair in order to defeat the PRD in his hometown of San Andrés Tuxtla, Veracruz (interview with Antemate). 192

201 had supported Cárdenas in 1988 fractured or wholly distanced themselves from the PRD in order to receive PRONASOL funds and other targeted PRI transfers. 45 Numerous scholars, including Molinar and Weldon (1994), Borjas (2003), and Magaloni (2006), have identified PRONASOL as a key determinant of the PRD s electoral setbacks in 1991 and Porfirio Muñoz Ledo echoed this assessment at the time. After the 1991 congressional elections, he wrote that through PRONASOL, the PRI had transformed from a party of the state a common characterization for decades into a party-state. 46 After the 1994 election, Muñoz Ledo, who had been elected party president a year earlier, assessed that the PRD had underestimated government agencies capacity for local manipulation through PRONASOL. 47 In the relatively small number of municipalities controlled by the early PRD, perredista mayors typically could not offset PRONASOL and other PRI programs with major projects and social policy initiatives of their own. Municipal governments in Mexico possessed paltry independent budgets and depended to a large degree on federal and state funds (Bruhn 1998: 229, 232). 48 Early PRD municipal governments received scant resources from the federal government and also, critically, 49 from state governments, which were under PRI and PAN control. Federal and state governments systematically postponed, rationed, and 45 A former Asamblea leader and PRD founder recalled, for example, that in Mexico City, the PRI, under the regency of Manuel Camacho, used social transfers to divide the Asamblea (interview with Rascón). 46 Proceso no. 776, 9/6/1991: For similar remarks by Cárdenas, see Proceso no. 773, 8/26/1991: Cited by Borjas 2003: Proceso, no. 931, September 5, 1994, pp In the USA, the federal government, in the mid-1990s, was responsible for less than 60 percent of total public spending, compared to roughly 90 percent in Mexico (Bruhn 1998: 227; Bailey 1994: 105). Due to fiscal centralization in Mexico, Bruhn (1998) writes that...to govern effectively[at the municipal level,] a party must successfully extract resources and cooperation from other state institutions and actors (Bruhn 1998: 227). 49 Magaloni (2006) shows that PAN mayors received more resources in states where the PAN also held the governorship (149). 193

202 sometimes flatly denied various forms of aid (not just PRONASOL resources) to PRDgoverned municipalities. 50 In summary, the PRI enjoyed access to vast public and private resources during the late 1980s and 1990s. With these resources, the PRI financed lavish campaigns and used targeted social spending to co-opt the left s natural constituencies. From the late 1980s until the second half of the 1990s, the PRD enjoyed virtually no access to state resources or finance. This resource disadvantage constituted a significant electoral handicap. The PRD s low access to media (late 1980s to mid-1990s) At the time of the PRD s emergence, most Mexicans depended on television to acquire political information, and even to form political views. Although print media had a significant impact on elite opinion, the vast majority of Mexican citizens did not read newspapers, given the country s high functional illiteracy rate and weak newspaper circulation. 51 As late as 2002, fewer than fifteen percent of Mexicans drew their news primarily from print media. 52 Radio played a similarly marginal role in providing political information. 53 Radio stations prioritized music and sports. Besides urban commuters and poor or isolated individuals without access to television, few tuned into the radio for news (Lawson 2002: 98). Control of the airwaves and, secondarily, print media conferred a major electoral advantage. Prior to the PRD s emergence, over decades of single-party rule, the PRI had thoroughly 50 On postponement and rationing, see Borjas (2003): 468. One party activist stated that PRI governors have always denied the PRD resources (interview with Córtez). Bruhn (1998) summarizes that the Salinas and Zedillo administrations undermined the PRD by controlling the flow of resources and shaping the party s ability to deliver on promises as a mediator for popular movements, as a local government, and so on (Bruhn 1998: 210, emphasis added). 51 Low education levels may also have made Mexican voters more susceptible to media influences than voters in developed democracies like the United States (Lawson 2002: 168). 52 Lawson (2002: 61); Vanden Heuvel and Dennis (1993: 40). 53 As of 2002, percent of Mexicans received political information from radio (Lawson 2002: 98). 194

203 penetrated national broadcast and print media through bribes, subsidies, penalties, and manipulation of broadcasting concessions. 54 These measures resulted in a docile, dependent, and captive media establishment that faithfully reflected the ruling party s priorities (Lawson 2002: 8, 173). In exchange for the government favoritism and support necessary for their economic prosperity and elite status (Vanden Heuvel and Dennis 1993: 21), the major owners of Mexican media, a few dozen in total, helped set the public agenda in broad accordance with PRI priorities, omitted sensitive, potentially damaging political developments from news programming, and broadcast open propaganda in support of the PRI and against PRI competitors (Lawson 2002: 8). At the core of Mexico s PRI-dominated media establishment lay the Televisa conglomerate (est. 1973). Ties between Televisa and the PRI ran deep. The Mexican television industry emerged in the early 1950s and took off in the 1970s, 1980s, and 1990s. From the 1950s onward, Televisa and its precursor, Telesistema Mexicano (est. 1953), received a succession of important concessions from PRI governments, including the country s original broadcasting licenses and, subsequently, licenses to broadcast cable, satellite, and high-definition television (Lawson 2002: 29). These concessions and other forms of government favoritism, such as generous subsidies for infrastructural development, enabled Televisa to consolidate monopoly status in the critical television sector. By the 1990s, Televisa commanded over four fifths of Mexico s television audience and received a similar share of television advertising revenue (Lawson 2002: 29). In exchange for government support, Televisa, like the media establishment more generally, tailored its news programming in a variety of ways to benefit the PRI. 55 In a telling quotation from the 54 According to Lawson (2002), before media liberalization, Mexico s media were thoroughly intertwined with the country s one-party system (8). 55 For more than two decades, Televisa and the PRI were deeply intertwined, with the network depending on the regime for concessions and infrastructure development, and the regime relying on Televisa for political marketing (Lawson 2002: 30). 195

204 mid-1990s, Televisa s chief executive, Emilio Azcárraga Jr., described himself as a soldier of the PRI and characterized the network as part of the governmental system (Lawson 2002: 30). 56 Predictably, throughout the 1990s, and especially during the first half-decade of its existence, the PRD suffered from low access to media and high levels of media hostility. The Salinas administration, through the Secretaria de Gobernación, used carrots and sticks, in particular the provision and cancelation of concessions and subsidies, to manipulate news coverage in detriment to the PRD. In order to gain or retain access to critical government support, media practiced self-censorship, omitting coverage of the PRD and denying the party advertising space. 57 Practitioners of self-censorship included major print media outlets. Before the rise of the center-right Reforma (est. 1993), the news weekly Proceso was the only significant medium in Mexico, print or broadcast, to consistently investigate and report on... closed topics and, more specifically, to provide a forum for left-wing political elites and intellectuals to communicate their views to a relatively large educated audience (Lawson 2002: 67). 58 Media manipulation mattered most, however, in the broadcast sectors. Mexico s major television outlets, led by Televisa, created and reinforced perceptions of left marginality by systematically blacklisting Cárdenas and the PRD. 59 A founding PRD elite and popular 56 PRI elites enjoyed similarly cozy relationships with subnational broadcasting outlets and major national newspapers including Excélsior, El Universal, El Sol, Novedades, El Diario de México, El Día, El Heraldo de México. Some of the PRI s informal media alliances (e.g., with the Heraldo de México) dated back to the Mexican Revolution (Lawson 2002: 29-30). 57 In addition, neither the FDN nor the early PRD could afford major television and radio spots (Bruhn 1998: 280). 58 In conversation, Borjas qualified this point, describing the readership of left publications like Proceso as a captive audience. She emphasized that they already supported the PRD. In her view, it was only through broadcast media that the PRD might have reached politically disengaged, undecided, and skeptical voters (interview with Borjas). 59 On occasion, broadcast station owners explicitly pressured rank-and-file not to discuss Cárdenas or report 196

205 movement leader stated that during early development, the PRD had no opportunities to disseminate a positive message via mass media. It was completely closed, he summarized. 60 Examples of PRD blacklisting abound. In the 1988 presidential campaign season, Cárdenas received under nine hours of airtime on Mexico s most watched cable news program, 24 Horas, while Salinas received over one-hundred and forty hours. 61 According to the Mexican Academy of Human Rights, Ernesto Zedillo (PRI), in the early months of the 1994 presidential campaign, received forty-six times more airtime than Cárdenas and Diego Fernández de Cevallos (PAN) combined. 62 Cárdenas repeatedly experienced the unforeseen cancelation of previously arranged media programs (Borjas 2003: 513), and during the NAFTA debates, the PRD s ability to affect public opinion was practically nonexistent due to informal media censorship (430-1). A party founder and ex-federal deputy recalled that in the smaller elections, early PRD candidates generally interviewed at independent radio stations with limited reach because the major networks, sold to the PRI, denied access to the left opposition. 63 In the rare events that PRD leaders received access to media, they often faced open hostility. 64 The paltry coverage of the PRD that media did provide highlighted the violent, subversive, and dangerous character of Cuauhtémoc Cárdenas and the party he led (Borjas 2003: 246). Throughout the PRD s formative phase, mass media touted the PRI 65 while depicting left critiques of the government. See, for example, Borjas (2003): 566). 60 Interview with Rascón. 61 Adler (1993): 155; Rodríguez (n.d.): See Bruhn (1998): 280; 63 Interview with Flores. 64 Borjas (2003) notes that television news hosts consistently displayed hostility and aggression toward perredismo and its leader during the 1994 campaign (513-4). 65 The counterpart of negative PRD coverage was positive PRI coverage (Lawson 2002: 52). While depicting the PRD as fringe and dangerous, media suppressed government criticism and cast the PRI as the party of security, prosperity, and public approval. In the 1988 presidential campaign, for example, Salinas was 197

206 Cárdenas and the PRD as a radical threat to social order. 66 During the 1988 presidential campaign, national television programs cast Cárdenas as the favorite of radical fringe groups, especially Marxist student organizations. In response to anti-fraud protests of , mass media and official propaganda depicted the PRD as a party of troublemakers who broke the law and disregarded institutions (Rodríguez n.d.: 288). 67 During the 1991 congressional campaigns, media characterized the PRD as violent and conflictual, ignoring the party s repeated condemnations of violent actions taken in the party s name. In 1994, media stoked public fears of violence and disorder 68 and attempted to resurrect the image of Cárdenas and the PRD as dangerous, zeroing in on Cárdenas s meeting with Subcomandante Marcos of the Zapatista National Liberation Army. Portrayals of the early PRD as dangerous and destabilizing extended to state and local elections as well. 69 One PRD scholar summed up the basic image of the PRD disseminated by television and other media outlets across the country: Being a perredista was something really, really bad. 70 always presented surrounded by multitudes (Rodríguez n.d.: 190), and Televisa launched a systematic campaign...to sell President Carlos Salinas program of economic reform (Lawson 2002: 101). During the implementation of PRONASOL, Televisa donated time for PRONASOL commercials, and its news programs gave priority to favorable coverage of the PRI (Bruhn 1998: 280). In 1993, Cárdenas stated that national media, due to pressure from the Salinas administration, were presenting an unrealistically positive image of Mexico to the United States (Proceso no. 847, 1/25/1993: 28). In the 1994 presidential campaign, there were twice as many positive PRI news stories as positive PRD news stories, and four times as many positive stories about Zedillo as about Cárdenas. Meanwhile, the PRI periodically censored critical media. For example, the Salinas government prohibited the circulation of allegedly defamatory publications such as Cambio 16 América, which accused Salinas of granting state favors to friends and family members (Borjas 2003: 435). 66 Borjas (2003) writes that mass media busied themselves with reprimanding the [early PRD s] behavior and disqualifying the party as a desirable governing alternative (Borjas 2003: 308). 67 According to Borjas (2003), coverage of marches, protests, and rallies deepened the belligerent and violent images of the PRD that the media emphasized (477). 68 During the election campaign, Televisa heavily covered violence in Guatemala, arguably with the intention to make voters more fearful, reactionary, and hence anti-prd (Lawson 2002: 53-4). 69 In the PRD s early bids for the governorship of Michoacán, for example, local newspaper coverage of the PRD was scarce and negative (Bruhn 1998: 236), and local media disseminated PRI slogans including Your vote decides: three years of violence or three years of peace, and Order and stability are the goals (Beltrán del Río 1993: 114). 70 Personal communication with Rodríguez, August 1,

207 As the left mobilizations of the late 1980s and early 1990s deintensified, news programs remained systematically critical of the PRD, but increasingly, instead of casting the PRD as a threat to social order, they structured their negative coverage around the party s policy stances and initiatives. Leading television programs generously broadcast and independently reinforced PRI critiques of the PRD on key issues such as NAFTA and the Zapatista movement. Describing the 1994 election, Borjas (2003) writes that the media, under the appearance of democracy, multiplied the airtime given to supporters of the Zedillo campaign so that they could discredit or criticize [the opposition], above all the PRD (566). Journalists often received negative feedback, and occasionally outright dismissal, if they attempted to cover party disputes in a balanced manner. During the NAFTA debates, for example, several media commentators lost their jobs after reporting that prominent intellectuals and regional politicians in Mexico supported the PRD s anti-nafta stance. 71 The combination of media hostility and low media access cost the PRD dearly in early elections. Gómez (1997: 16) attributes the PRD s initial electoral weakness to two primary factors: PRI repression and, above all, the media s perpetual smear campaign (campaña de desprestigio). Similarly, Borjas (2003) identifies Mexican media s hostility and informal censorship as a central determinant of the PRD s electoral setbacks in 1991 and 1994: With the help of the media in 1991, the PRI recovered and maintained [the initiative] through the orchestration of a permanent campaign against the PRD, and in 1994, the fraud did not take place in the voting booths, but was a process fundamentally based on inequality of resources and access to media and the [media s]...manipulation of information (514-5, 571, 587) This story was picked up by the New York Times, which increased international attention to press censorship. As a result, Mexico s Director of Radio, Television, and Cinematography, in charge of private media regulation, resigned (Borjas 2003: 429). 72 See also Borjas 2003:

208 Mexican media started to become more competitive and state-independent in the early to mid-1990s. Televisión Azteca was born in 1993 and, over the course of the decade, became consolidated as Televisa s main competitor. The 1994 election included Mexico s first ever televised presidential debate, featuring the PRI s Ernesto Zedillo, Cárdenas, and the PAN s Diego Fernández de Cevallos. Cárdenas began his opening remarks with the statement: This opening of the media is, without a doubt, without any doubt, an achievement for those of us who, from distinct positions, from different organizations, are struggling for democratic change in our country. 73 Yet media liberalization during the early to mid-1990s should not be overstated. The initial process of media opening occurred slowly and cautiously. Mexico s major television news programs continued to provide circumscribed, unassertive reporting past the mid-1990s. Televisa s political coverage remained tentative and halting, and as late as 1996, only a small fraction [of its news coverage]...was devoted to... closed topics such as official corruption, drug trafficking, electoral fraud, anti-government protests, and the Mexican military (Lawson 2002: 105). TV Azteca (est. 1993) did not provide authentic news coverage during its early years, focusing primarily on entertainment (Lawson 2002: 104). Independent print media did not begin to take off until the mid-1990s, with the rise of Reforma, El Financiero, and La Jornada. Among these, only La Jornada was located on the left of the ideological spectrum. 74 * * * Due to the four adverse conditions identified in this section fraud, repression, resource disadvantage, and media blacklisting and hostility the early PRD did not win any significant positions of power. In contrast to the PAN, the PRD did not win a single governorship or See Lawson (2002): 62-9 for a discussion of print media liberalization. 200

209 major mayoralty (save Morelia, Michoacán) before the late 1990s. 75 In Mexico s centralized, authoritarian political system, small mayoralties and minority legislative blocs at the state and federal levels, through the mid-1990s, did not give the PRD policy autonomy, control of large government agencies, or significant public financing. One party activist and scholar summarized in an interview that past the midpoint of the 1990s, there was nothing to distribute : no patronage, no salaried positions. 76 Media and state access during the late 1990s In several important respects, the political landscape, and the PRD s fortunes, began to change toward the end of the 1990s. The landmark electoral reforms of November 1996, which the PRD played a critical role in devising and legislating, constituted a major step in Mexico s democratic transition. The reforms made the critical Federal Electoral Institute an independent body, opened the Federal District to direct elections, 77 placed restrictions on private campaign spending, 78 and provided all registered political parties, including opposition parties, with increased access to public television and much more generous public financing. 79 The reforms also increased the requirements for party registry in order to prevent tiny parties from cluttering the national political system (Klesner 1997). Also toward the end of the 1990s, Mexican mass media began to provide more balanced political coverage (Lawson 2002; Navarrete 2010: 264). In the 1997 congressional elections, opposition parties were finally given a real chance to present their views (Lawson 2010: 109). Cárdenas stated in Proceso that the unquestionable opening of the media made 75 The PAN won its first governorship (Guanajuato) in Interview with Borjas. 77 Previously, the president appointed the regente of the Federal District. 78 Only ten percent of campaign spending could come from private sources. 79 Major-Party Deal in Mexico to Bring Political Reforms, New York Times, July 27,

210 his 1997 victory in the Federal District possible. 80 Also in 1997, the PRI experienced an unprecedented decline in popularity among Televisa viewers (Lawson 2002: 170). By the late 1990s, Televisa had evolved from a private Ministry of Truth into a more typical commercial network (Lawson 2002: 10). Yet even the media liberalization of the late 1990s should not be overstated. Lawson (2002) observes that media network owners, cross-nationally but especially in Mexico, tend to hold conservative views and tilt coverage in favor of conservatives: Nowhere has this been clearer than in Mexico, where television coverage has been substantially more sympathetic to the PAN than to the PRD and remains constrained on any topics that might be construed as leftist (Lawson 2002: 208). Thus, in 2000, most mass media strongly opposed Cárdenas. 81 To the present, perceptions of media bias remain widespread within the PRD. 82 Partially as a result of the above developments the 1996 reforms in finance and public media, the liberalization of private media the PRD, in the late 1990s, began to invest more heavily in mass media initiatives and win more major elections. The new political party fund divided all registered parties yearly allotment into two categories: ordinary expenses and campaign expenses. In 1997, the first year in which registered parties received the enlarged party fund, the PRD s CEN allocated seventy percent of its campaign expenses to television and radio advertisements and paid media spots (Borjas 2003, vol. II: 124). In the years 1997 to 1999, the PRD won its first major executive victories. In 1997, Cárdenas prevailed in the first ever election for jefe de gobierno of the Federal District, the second most important 80 Proceso, no. 1067, April 13, Authors interview with Córtez. 82 At the federal and subnational levels, perredista activists and elites repeatedly complain of unbalanced coverage on the part of private media, especially Televisa. Such complaints proliferated and intensified after the 2006 and 2012 presidential elections, during which PRD supporters alleged systematic smear campaigns against Andrés Manuel López Obrador. The Yo Soy 132 social movement, created in the lead-up to the 2012 election, lists as its central goal the democratization of Mexican politics and media. 202

211 elected office in the country. In 1998 and 1999, the PRD won its first three governorships in Zacatecas, Tlaxcala, and Baja California del Sur. Access to mobilizing structures during the PRD s formative phase The PRD s early adversity shaped elite electoral strategies. With very little access to state resources, finance, and media, the PRD leadership had limited options for building electoral support. Television, radio, and most newspapers cast the party in a negative light and denied party leaders meaningful opportunities to counter this image. The party controlled no important governments just a small number of resource-strapped municipalities and thus could not build a reputation, even subnationally, for effective governance. Given the party s lack of access to state patronage, public finance, and private finance, candidates could not run professional campaigns with teams of consultants and strategists and armies of electoral workers, all incentivized by salaries or promises of government spoils. The PRD would rise or fall, in large measure, on the strength of its volunteer ground organization. PRD founders understood this. Resource asymmetries, media ostracism, and media disinformation all rankled the party elite. Proceso featured numerous exposés of the Mexican television industry, often written by PRD intellectuals and elites, which focused on the major networks right-wing bias and hostility toward the left opposition. 83 Candidate speeches and campaign pamphlets treated local PRD activists as essential to the party s success: only the militancia, these materials emphasized, could neutralize the negative influence of media by delivering the PRD s unfiltered message to voters. 84 Four party founders emphasized, in interviews with the author, that media blacklisting and PRI political hegemony made 83 See Borjas (2003): 246 note Examples needed. 203

212 the PRD s ground campaigners fundamental. 85 Borjas (2003) finds that in the early 1990s, penury and media ostracism dictated PRD strategy, impelling party candidates to base their campaigns on exhaustive local tours and large rallies in town plazas (486). 86 Elite incentives, however, were not sufficient for the emergence of a strong PRD organization. Low access to media and state resources gave PRD elites an incentive but not the capacity to acquire activist networks and organizational infrastructure in large swathes of the national territory. Lacking resources, the central PRD office could not finance or administer the formation of thousands of local party branches. Early organizational expansion could only occur from the bottom up, through the initiative, and with the dispersed resources, of local civil society leaders and the organized networks under their control. Access to preexisting civil society structures and opposition party networks gave PRD elites the capacity, or means, for organizational implantation. On creation, the PRD embarked on a mission of large-scale, base-level expansion, seeking to establish local committees in as much of the national territory as possible. 87 Acutely aware of the national party s resource constraints, the PRD leadership set out in 1989 to forge linkages with leaders from across Mexican civil society; in Porfirio Muñoz Ledo s words, to base the party on what existed. 88 With contact and encouragement but little aid or oversight from the PRD s central office, local leaders took it upon themselves to organize local PRD branches and nuclei. In a rejection of PRI-style corporatism, the PRD prohibited feeder organizations from transferring their members to the PRD as a single bloc, at least formally. 85 Interviews with Rascón, Saucedo, Gamundi, and Hidalgo. 86 The insufficiency of economic resources limited the penetration of PRD campaigns, given that, in addition to being ostracized by the media, the PRD could not buy advertising space... Under these circumstances, the PRD s electoral strategy [in the early 1990s] remained based on direct contact with the people through candidates exhaustive local tours and the organization of rallies in all the main plazas (Borjas 2003: 486). 87 Interview with Gamundi. 88 Interview with Muñoz Ledo. 204

213 The PRD sought to become a party of citizens based on individual membership...in which the axis of party life was participation in the comité de base (Rodríguez n.d.: 254). The local leaders who built the PRD organization at the ground level came primarily from three groups: the traditional Marxist left, the social left, and ex-pri networks. The traditional Marxist left provided a small activist base but a large reserve of educated, experienced political elites and cuadros. The social left provided a large social base, divided into numerous local activist networks and grassroots structures in most of the PRD s early strongholds, including the Federal District and southern states. In the CD s territorial bastions of Michoacán and Tabasco, the PRD inherited significant core electorates, large activist networks, and valuable organizational machinery from the PRI. The traditional Marxist left For most of the twentieth century, the Mexican Communist Party (PCM, est. 1919) dominated Mexico s Marxist landscape. Yet in the late 1960s, due to the background influence of the Cuban Revolution and the galvanizing effects of the Díaz Ordaz (PRI) government s violent repression of UNAM students, new Marxist parties, organizations, urban and rural guerrilla groups, and intellectual cliques proliferated. Some were PCM splinter groups. Both the traditional Marxist parties and most of the social left organizations that would eventually feed into the PRD emerged from this period of new Marxist ferment in Mexico. During the 1970s and 1980s, as a consequence of diminishing repression and increasing political liberalization, much of Mexico s new Marxist left moderated. These groups ended or reduced their participation in extraparliamentary forms of struggle, including guerrilla warfare, and created political parties for the purposes of electoral participation. Heberto Castillo, later to become the most important Marxist founder of the PRD, created the Mexican Workers Party (PMT) in 1974, after spending two years in prison from 1969 to 205

214 1971. Other important parliamentary Marxist parties founded between the mid-1970s and early 1980s include the Trotskyist Socialist Workers Party (PST, est. 1975), the Trotskyist Revolutionary Workers Party (PRT, 1976), and the Revolutionary Patriotic Party (PPR, est. early 1980s). The PPR was an insurgent successor party, founded by members of the urban Liga Comunista 23 de Septiembre, Mexico s strongest guerrilla organization. 89 These actors, together with the PCM, came to compose the traditional Marxist left. From the mid-1970s to the late 1980s, the parties and groups of the parliamentary Marxist left, separated for decades due to rigid sectarian divisions, forged a succession of electoral alliances. First, in 1976, the still-proscribed PCM joined with a group of smaller parties to form the Left Coalition (Coalición de Izquierda, or CI), an alliance backing the 1976 presidential bid of the PCM s Valentín Campa, a railway union leader and former political prisoner. Salazar received 600,000 votes, registered as invalid due to the legal ban on left opposition parties. The growing strength of the organized left opposition led the PRI regime, in 1977, to lift the legal ban on left opposition parties. The PCM and PST quickly obtained registries, and several small parties united with the PCM to contest the 1979 midterms. The PCM won 703,000 votes, consolidating its status as Mexico s third electoral force. The push for left electoral unity continued in the 1980s. Key leaders of the traditional left judged, early in the decade, that a more united Mexican left could prove a true counterweight to the PRI: It was calculated that electoral-political unity would lift the left from marginality and could convert the left into a real protagonist in the country s decisionmaking (Hérnandez 2010: 33). Thus, in late 1981, the PCM, in its twentieth Congress, dissolved itself and transferred its registry to a new, larger left party, the Mexican Unified Socialist Party (PSUM). PSUM leaders chose March for Democracy (Marcha por la Democracia) as the presidential campaign slogan for Arnoldo Martínez Verdugo, who re- 89 The Liga claimed nuclei in all of Mexico s major cities. 206

215 ceived nearly four percent of the vote. The creation of the PSUM marked an important step in the unification of Mexico s traditional Marxist left, but the process would culminate in the formation of the PMS. Like the 1982 presidential election, the 1988 election gave Mexico s still-fragmented traditional Marxist left an incentive to seek even greater unity. In 1988, the PSUM dissolved itself and transferred its registry to the new PMS, which absorbed, in addition to the PSUM, six Marxist parties and organizations: Heberto Castillo s PMT, the PRT, a defecting faction of the PST led by Jesús Ortega, the Union of the Communist Left (UIC), and the more radical Revolutionary People s Movement (MRP, est. mid-1960s) and aforementioned PPR (Hérnandez 2010: 36). The creation of the PMS marked the culmination of a decades-long process of left unification. When the PMS nominated Heberto Castillo as its candidate for the 1988 presidential election, the party claimed a broader coalition of left forces than any left party in Mexican history. As disussed earlier, however, Cárdenas s independent 1988 presidential bid rapidly eclipsed Castillo s and threatened the PMS s survival. Consequently, the PMS retired Castillo s candidacy, joined the FDN, and in 1989, dissolved itself and transferred its registry to the PRD. The PRD thus inherited the PMS wholesale. The PMS claimed party representatives and local or regional branches across the national territory. 90 On adopting the PMS s registry, the PRD inherited several thousand PMS members and activists, a network of offices (particularly in large urban areas), and additional organizational resources (e.g., telephones). Porfirio Muñoz Ledo, the PRD s first national Secretary of Organization, recalled that during the late 1980s, the national PRD leadership contacted or visited PMS offices across the country in order to develop an initial territorial network of local PRD nuclei. 91 The PMS s national headquarters, located at Monterrey Interview with Gómez Tagle. 91 Interview with Muñoz Ledo. 207

216 in the Federal District, became the PRD s first national headquarters. Yet the PMS s role in the PRD s territorial implantation should not be overstated. The PMS s most critical contribution lay in the provision of experienced, educated elites and cuadros. A disproportionate number of these individuals rose to positions of national leadership and won major offices. The party s several thousand members, however, did not represent a significant share of the PRD s mass base. As Cuauhtémoc Cárdenas observed in a 2001 interview: I would say that the PMS is an organization that comes with many cuadros but without a broad popular base in general terms. 92 To build a mass base, the PRD depended much more heavily on the extraparliamentary social left and, to a lesser extent, the defecting PRI. The social left Following the PRI s brutal repression of students in 1968, many left organizations, in contrast to the above Marxist parties, continued to reject institutional politics and engage in social struggle throughout the 1970s and 1980s. These organizations disdained elections, detested the PRI, claimed férreas militancias (iron-willed activist bases), and engaged in warlike (including armed) forms of struggle (Martínez 2005: 53, 55). Many included individuals who had split from traditional Marxist parties or taken up parallel activities after The social left organizations that fed into the PRD, almost all of which had actively supported the FDN in the 1988 presidential campaign, either possessed large social bases or, more commonly, strong ties to organizations with large social bases. Key social left feeder organizations with large bases included the National Worker and Peasant Council of Mexico (CNOCM); the Democratic Peasant Union (UCD); the University Student Council; Interview with Cuauhtémoc Cárdenas in Borjas (2003, vol. II: 410-1). 93 Hernández (2010):

217 the Worker, Peasant, and Student Coalition of the Istmo (COCEI, est. 1976); the Popular Urban Movement (est. 1985) and, within it, the Asamblea de Barrios (est. 1987); the University Student Council (CEU, est. 1986); and the peasant-based Cardenista National Council (CNC). Those with with strong ties to large social bases included parties and cliques such as the Organización Revolucionaria Punto Crítico (ORPC, est. 1972), a Mexico City-based Marxist organization founded by student movement leaders and in charge of the influential left magazine, Revista Punto Crítico; the Popular Action Movement (MAP, est. 1980), an ORPC splinter; the Línea de Masas faction of the Maoist Revolutionary Leftist Organization (OIR- LM, est. 1981); the Revolutionary People s Movement (MRP, est. 1981); the Maoist Revolutionary National Civic Association (ACNR, est. 1982), a nucleus of former guerrillas; the Revolutionary Patriotic Party (PPR, est. early 1980s), also a member of the PMS; the Green Party (PV, est. 1986); the Movement Toward Socialism (est. March 1988), a PRT splinter; the Grupo Poliforum; the Liberal Party; and Democratic Convergence. 94 These cliques and small parties claimed strong ties to a range of organized social sectors, including peasants (ACNR), urban popular movements (ORPC), student associations in Mexico City, Guerrero, 95 and elsewhere (MAS, ACNR), left teachers unions such as STUNAM and the CNTE (MAP, OIR-LM), left blue-collar unions (MAP), and post-materialist movements (PV). Although a wide range of social left actors contributed to the organizational development of the PRD, rural unions and urban popular movements played the most important role in creating permanent PRD base-level structures (Rodríguez n.d.: 257-8). Peasant unions played an especially critical role in southern Mexico. Rural worker associations provided ex- 94 See Martínez (2005: 53-5); Hernández (2010: 37-8); Rodríguez (n.d.: 188); Borjas (2003: 279). 95 One PRD founder and early activist notes the key role that left student groups at the Autonomous University of Guerrero played in early PRD mobilization and organization (interview with Gamundi). 209

218 tensive activist networks and infrastructure. The aforementioned CNOCM, a confederation of progressive rural and urban unions that had split from the Mexican Workers Confederation to support Cárdenas s 1988 presidential bid, 96 gave the PRD an organized rural base in Michoacán, Guerrero, Oaxaca, Chiapas, Tabasco, and Veracruz. According to Bruhn (1998), local branches of the Democratic Peasant Union (UCD) simply became local PRD committees in many cases: Membership of local PRD/UCD committees often overlapped nearly 100 percent (225). One PRD founder described rural unions as the early PRD s spinal cord in large parts of the southern peasant states. 97 The Popular Urban Movement (MUP) and especially the Asamblea de Barrios, which the MUP produced played an indispensable role in the early PRD s organization-building in Mexico City. The MUP (est. 1985) emerged in response to Mexico City s devastating 1985 earthquake. The earthquake wrought havoc on the city s vast popular neighborhoods, destroying homes and the supporting infrastructural systems. Widespread dissatisfaction with the pace and scope of the de la Madrid administration s response sparked the MUP, which involved an unprecedented flowering of neighborhood associations dedicated to housing rights advocacy and community reconstruction. Hérnandez (2010) identifies the 1985 earthquake as the most significant event for the development of Mexican civil society that had occurred in decades (35). With the rise of the MUP, popular urban sectors, traditionally a PRI constituency, defected to the opposition. Porfirio Muñoz Ledo, in a 2009 interview, summarized that the citizenry, confronting the inadequacy of the de la Madrid administration, organized itself, and the government s authority in [Mexico City] was over. 98 The city s proliferating neighborhood 96 Cárdenas (2010). 97 Interview with Gamundi años del PRD: Porfirio Muñoz Ledo, May 22,

219 associations united to form numerous umbrella organizations. 99 In 1987, the largest of these, the Coordinadora Única de Damnificados, dissolved into an even larger confederation, the Asamblea de Barrios. 100 Months after Cárdenas announced his candidacy, the Asamblea endorsed him. Other important Cárdenas allies from the MUP included the Unión de Colonias Populares and the Frente Popular Francisco Villa. These organizations fed into the PRD on its founding. In the Federal District, the PRD inherited a mass activist base and dense grassroots structures from the MUP. In the late 1980s, the Asamblea alone claimed 60,000 member families, an estimated eighty-five percent of whom became perredistas. 101 One PRD founder and early federal deputy summarized that the MUP enabled the PRD to move beyond university students and radical unions and become rooted in mass society. 102 neighborhood activists and groups into local PRD committees. 103 MUP leaders mobilized Many became PRD candidates for public office. 104 The Asamblea lent Cárdenas and the PRD key assets, including panel trucks for campaigning. 105 Such was the Asamblea s contribution to Cárdenas and the 99 These included the Unión de Colonias Populares, the Frente Popular Francisco Villa, the Unión Popular Nueva Tenochtitlán, in which future PRD leaders René Bejarano and Dolores Padierna participated), the Coordinadora de Cuartos de Azotea de Tlatelolco, the Comité de lucha Inquilinaria del Centro, the Unión de Vecinos de la colonia Pensil, and the Unión Popular de Inquilinos de la colonia Morelos-Peña Morelos. 100 See Hernández 2010: Interview with Rascón. 102 Interview with Ávila. 103 Local PRD committees, Bruhn (1998) writes, tended to coalesce around preexisting pockets of movement organization, particularly those connected to the Asamblea (226). A PRD founder and early federal deputy recalls that in the Federal District, PRD activist groups were transplanted from urban popular movements (interview with Flores). Another PRD founder early federal deputy states, similarly, that he and other urban popular movement leaders mobilized PRD support in their movements neighborhood strongholds (interview with Saucedo). Both interviewees drew on their recognition and support among the movement base to build large, local constituencies and launch successful campaigns for federal deputy (interview with Flores). 104 Successful federal deputy candidates drawn from urban popular movements included Marco Rascón and Francisco Saucedo of the Asamblea, Cuauhtémoc Memo Flores of the Coordinadora única de Damnificados, Alejandro Luebano of COPEDI and the Unión de Colonias Populares, and René Bejarano of the Unión Popular Nuevo Tenochtitlán. 105 When campaigning in Mexico City in 1988, Cárdenas made many of his appearances in the Barrio Móvil, 211

220 PRD that, in the words of one movement leader, the Asamblea destroyed (deshizo) the PRI in the Federal District. 106 Mexico s social left movements and organizations, taken together, provided the PRD with a larger social base than any other feeder category. In addition, the social left furnished a large number of experienced leaders and cuadros who occupied important positions in national and subnational party organs. These individuals exerted a profound influence on activist formation and discussions of party program and tactics. Early party meetings, for example, predominantly consisted of radical activists and intellectuals. 107 One ex-member of the ORPC recalls that groups like Punto Crítico played a critical role in educating PRD activists and leading discussions of political theory and coyunturas. These contributions, he stated, fueled the political and ideological struggle. 108 It should also be noted, finally, that many perredistas of extraparliamentary Marxist origin viewed the PRD as a stage, important but temporary, in a longer-term struggle for revolutionary, socialist change in Mexico. 109 A significant fraction of these individuals, after joining the PRD, maintained parallel, clandestine activities, especially until the mid-1990s. Ex-PRI networks During the PRD s early years, ex-pri networks furnished key human and organizational resources in several states, especially Michoacán and Tabasco. 110 Michoacán constituted a large panel truck that the Asamblea had created primarily for public presentations of their mascot, Super Barrio Gómez (personal communication with Greene, June 10, 2011). 106 Interview with Rascón. 107 Personal communication with Bruhn, June 10, Interview with Gamundi. 109 For example, one early PRD federal deputy, a Marxist and MUP member, recalled that when he joined the party, he believed that revolution would ultimately prove necessary, but that achieving electoral success would help the left acumular fuerzas, or amass power (interview with Flores). 110 It should be noted that many defecting priístas simultaneously participated in social left organizations (e.g., peasant unions), but most did not. 212

221 the primary bastion of cardenismo and perredismo throughout the formative phase. Lázaro Cárdenas served as the governor of Michoacán from 1928 to 1930, and Cuauhtémoc Cárdenas served in the same position from 1980 to Prior to the creation of the FDN and PRD, a large fraction of Michoacán s electorate identified as cardenista, not priísta. 111 Thus, when Cárdenas defected, most of the PRI organization and electorate defected with him. In the 1988 election, Cárdenas received the majority of the state s vote. On formation, the PRD in Michoacán inherited a large network of ex-pri cardenista supporters, activists, and cuadros. In a 1992 edition of Proceso, Porfirio Muñoz Ledo estimated that in Michoacán, 98 percent of our militantes and cuadros were from the PRI. 112 Of the three-hundred mid-level leaders (cuadros) who defected from the PRI 1988, two-hundred and fifty hailed from Michoacán (De la Madrid 2004). Fernández (2003) states that the PRD of Michoacán came whole and ready-made (en forma integra) from the PRI. The PRD also inherited valuable organizational assets such as offices, computers, and telephones. 113 Ex-PRI networks also played a key role in the PRD s early implantation in Tabasco. Beginning in 1989, the first president of the PRD s Tabasco office, ex-priísta and CD member Andrés Manuel López Obrador (AMLO), built a statewide network of precinct-level PRD committees (comités de sección). He based these committees on a similar network of PRI committees that, as Tabasco s PRI president in 1983, he had helped develop and integrate horizontally. In 1993 and 1994, AMLO, in anticipation of the 1994 gubernatorial race, undertook a second major organizational push, to impressive effect. Building on his initial network of local PRD comités, AMLO spearheaded a new, major recruitment effort, resulting in a surge of PRD members and local nuclei. By 1995, the PRD claimed a higher 111 Interview with Flores. 112 Proceso, no. 820, July 20, According to Borjas s (2003) less precise estimate, the majority of early PRD activists in Michoacán were ex-priístas (355). 113 In many parts of Michoacán, according to one PRD founder, the PRI organization simply became the PRD organization, in terms of personnel as well as resources (e.g. computers) (interview with Flores). 213

222 member/population ratio in Tabasco than in any other Mexican state. 114 * * * In sum, while low access to state and media made territorial implantation electorally vital for Mexico s new left, access to human and organizational resources from the traditional Marxist left, social left, and former PRI made it possible: The credit for the PRD s territorial implantation fundamentally goes to the work carried out by leaders and members of the local organizations and movements that were incorporated into the party...[and] contributed human and material resources to party implantation (Borjas 2003: 245). 115 The early PRD s organizational strength Within several years of its creation, the PRD had built a territorial infrastructure that, in reach and density, rivaled or surpassed that of nearly every new left party in Latin America on a comparable timeline, including Brazil s PT. 116 The party claimed roughly 700,000 members in late 1991 and, by 1995, over one-million. 117 Geographical unevenness followed from the diffuse, bottom-up nature of the early PRD s organization-building. In the territorial bastions of the PRD s organized allies, PRD branches proliferated. Elsewhere, few branches materialized. 118 Early perredistas heavily populated 114 Borjas (2003: 519); Borajs (2003, vol. II: 74). 115 Similarly, an early activist recalls that the PRD became territorially rooted through locally implanted organizations, especially where big leaders of social movements operated (interview with Ávila). 116 Although the PRD, in regional comparative perspective, has a high level of territorial rootedness, PRD members and analysts tend to downplay or deny the party s infrastructural strength. In interviews, dozens of party elites and activists described the party as infrastructurally weak. This assessment usually resulted from comparing the PRD to the PRI, which has penetrated every region of Mexico. 117 First statistic from Borjas (2003: 371). 118 One PRD leader and federal deputy states that wherever the early PRD received significant support, it came from already-organized left groups (interview with Flores). Borjas (2003) writes of the PRD s 214

223 the Federal District, Mexico state, and much of the South, but not the North or, for the most part, Central Mexico. At the PRD s first two Congresses, the largest delegations came from the Federal District and Mexico State, Michoacán, Veracruz, and Tabasco. 119 By 1993, the PRD had set up a party office in every municipality of Michoacán and established a significant presence, proportional to population, in six additional states: Chiapas, Guerrero, Oaxaca, Hidalgo, Tlaxcala, and Zacatecas. 120 Throughout this period, the PRD claimed a strong, active base in the Federal District and, to a lesser extent, Mexico state, but given high population size in both entidades federativas, the party s proportional presence was comparatively weak. In 1994, due to electoral mobilization for the presidential and gubernatorial contests, the PRD s percent membership in the Federal District and Tabasco skyrocketed. As of 1995, the party s member/population ratio in these two entidades surpassed those in all other Mexican states, 121 and the majority of PRD members, nationally, resided in six entidades: the Federal District, Michoacán, Tabasco, Veracruz, Oaxaca, and Guerrero. 122 early organization-building process: The diversity of social movements and organizations that joined the PRD...in some places, led to an important and almost immediate territorial implantation derived from the rootedness (arraigo) and mobilizing capacity of those who headed them and formed part of the PRD s organizing committees... (301). 119 Borjas (2003): 371 fn. 284, The party had a weaker, but moderate, presence in Aguascalientes, Baja California, the Federal District, Jalisco, Mexico state, and Nuevo León (Borjas 2003: 518). 121 See Borjas 2003, vol II: 74). 122 Borjas (2003, vol. II): 74,

224 Figure 4.2: PRD Municipal Office Presence, 1990 Figure 4.3: PRD Municipal Office Presence,

225 Figure 4.4: PRD Member/Population Ratio, 1995 Polarization and the sources of activist commitment In sum, the PRD, during its first half-decade, built a strong organization with over 1000 municipal offices, roughly one-million members, and a mass activist base, concentrated primarily in the Federal District, Mexico state, and the Southern region. PRD activists, the lifeblood of this organization, possessed high levels of partisan commitment, enthusiasm, and drive. In unprompted recollections, various PRD founders vividly attested to this fact, describing the early militancia as a mass of enthusiastic volunteers characterized by idealism, a willingness to work around other important responsibilites, a knack for creative grassroots mobilization, and an aversion to central features of modern Mexican campaigning such as patronage, clientelism, paid media spots, and political professionalism. Distilling a commonly expressed sentiment, one PRD founder and ORPC member stated 217

226 that early PRD activists enjoyed party work. 123 Another party founder and MUP member stated that for most early PRD militantes, political activism was natural and the concept of political professionalization almost non-existent. More clinically, one party activist and scholar stated that the psychological benefits of participation drove the early PRD s big volunteer army. 124 A founding PRD member, formerly of the PSM, described the PRD s convocation as a great outpouring of spontaneous, vigorous debate and analysis among thousands of members and numerous small parties and organizations. 125 In early speeches, party elites frequently trumpeted the PRD s spirit of volunteerism and contrasted the PRD s creative and spirited grassroots campaigns with the money-fueled, professional campaigns of the PRI. One interviewee fondly recalled early PRD propaganda as highly artisanal (artesanal) and stated that creative, low-budget propaganda had more credibility than a [expletive] TV ad. 126 A PRD founder and intellectual stated, without prompting, that the party s mass reserve of enthusiastic, spontaneous volunteer activism constituted its core strength and comparative advantage. For no other party during the early 1990s, he said, did small groups of individuals routinely engage in spontaneous grassroots activities such as writing PRD slogans on asphalt sidewalks. 127 One early PRD activist noted the absence of bureaucracy and structure during early ground campaigns and also highlighted the untutored quality of his and others early party work. He and fellow activists, he said, allocated time outside of their work and study schedules 128 and improvised with little supervision, campaigning door-to-door, 123 Interview with Gamundi. 124 Interview with Borjas. 125 Interview with Ávila. 126 Interview with Rascón. 127 Interview with Gamundi. 128 Along these lines, one PRD founder and future federal deputy from Mexico state, in a 2010 interview, described doing party work in the evening, after classes: When I began as a PRD activist, I studied at 218

227 painting graffiti on walls, and distributing locally produced fliers. He recalled that many such individuals, including he, continued their party work between election campaigns, with a focus during these periods on forging ties to popular organizations. All these efforts, he stated, were spontaneous and decentralized, with each activist aware that s/he needed to make a contribution to building the PRD. 129 One interviewee emphasized that during the artisanal period, early PRD militantes typically invested at least some of their energy in local elections. 130 Since effective ground campaigns conferred a potentially decisive advantage in smaller elections, he stated, many activists assigned top priority to local contests and devoted most of their party work to local candidates. These individuals did not expect the PRD to win national power easily. On the contrary, they expected, even welcomed, a gradual, bottom-up transformation of Mexico s political system, in which the left would amass victories and governing experience at the local and state levels before winning control of the federal government. 131 In 1991, the national leadership followed the base s lead by establishing its first set of official principles and guidelines for PRD municipal governments. 132 What factors account for the commitment and enthusiasm of the PRD base? Selection pressures played an important role. The PRD s early adversity, in addition to shaping elite the Colegio de Ciencias y Humanidades during the afternoon and evening. I would get home at 9. At 10, with my brother, I would hang up posters and distribute fliers (Interview with Duarte in González et al., eds. (2010: 250). 129 Interview with Hernández. 130 Interview with Rascón. 131 Interview with Rascón. 132 Initially, the PRD possessed clear national objectives (e.g., democratization, defeating the PRI, rolling back neoliberalism), but the party did not consider an initial proposal for winning state and local governments. In 1991, the national party, aware of the importance of winning positions at diverse levels of government, organized a national municipal assembly and produced a multipoint set of guidelines for PRD municipal governments (Baptista 2010: 175). Six more assemblies would follow, and in 1997, the national PRD would create two organs, the Institute of Municipal Development (IDM) and the Mexican Association of Local Authorities (AALMAC), to support its municipal governments (Baptista 2010: 175-6). 219

228 electoral strategies, created selection pressures at the base level. Active participation in the early PRD presupposed a willingness to do party work without a salary or the credible promise of a government job. The party s first national Secretary of Organization, Porfirio Muñoz Ledo, deftly noted in an early Proceso article that we live by substituting good faith and enthusiasm for resources. 133 Bruhn (1998) notes that most early perredistas also paid less visible costs, having to take time out of profitable work activities and expend income on fuel, food, and travel lodging (286). For many, participation in the PRD required a willingness to put one s own safety at risk, especially in the more violent southern states. Rodríguez (n.d.) thus writes that being a victim of persecution and exclusion, experiencing electoral fraud, even losing one s life were constant possibilities for party activists (284, emphasis added). Unremunerated labor, dim short-term electoral prospects, and the threat of violence and murder discouraged political careerists (arribistas), opportunists, and patronage-seekers from joining the early PRD. The PRD s adverse formative phase thus selected for militantes creyentes, committed activists who sincerely believed in the PRD s causes. Greene (2007) finds evidence of these selection effects. Individuals who joined the early PRD in years of heightened resource asymmetry and repression, he shows, held more extreme left-wing views on economic policy and democracy, holding constant economic development, a host of demographic variables, and activists formal position in the PRD hierarchy (158, 305). 134 One party activist and scholar stated in an interview that conditions during the PRD s early development were very difficult, but that this difficulty helped the party. Because the early PRD had very few opportunities to win important elections, she stated, the party couldn t 133 Proceso, no. 757, May 6, 1991, p. 19. Cited by Borjas (2003): In particular, many of the PRD s early ideologues came from Mexico s older, radical left (Greene 2007: 164) and had themselves been recruited into the opposition when joining was such a high cost and low benefit activity that only relative extremists found it worthwhile to do so (94-5). 220

229 be a business. Perredistas joined for other reasons, like ideological conviction. 135 Borjas (2003) also suggests the existence of selection effects, writing that the repression exercised by the government led...to the despondency and desertion of the FDN s base in many cases, but not for the most strong-willed individual and group supporters (341, 245). Selection pressures alone, however, cannot account for the commitment of early PRD activists. Even if a particular set of conditions selects for committed activists, individuals and groups will not donate their time, energy, and resources to a party-building project unless driven by a positive higher cause. Party-building commitment on a national scale presupposes a higher cause of national import, around which large, heterogeneous, and dispersed segments of a population can unite and rally. Independently of selection pressures, such a cause, in a given context, may or may not emerge. Steven Levitsky and his collaborators have argued that periods of national polarization and conflict often produce the higher causes necessary for large-scale, volunteer collective action and party-building (Levitsky et al. n.d.). The early PRD conforms to this pattern. Intense polarization and conflict between Mexico s governing powers and left opposition during the late 1980s and early 1990s generated the common causes that fueled early PRD activists. Cardenismo, the broad movement that fed into the PRD, both reflected and contributed to this polarization. By the late 1980s, the PRI, governing amid economic and humanitarian crisis and failing to respond adequately, could no longer legitimate its rule by citing strong performance. In increasing numbers, Mexican voters and civil society organizations began to defect from the PRI and join the opposition. The marginalization of the Corriente Democrático, headed by Cuauhtémoc Cárdenas, revealed the PRI s internal resistance to change. When Cárdenas defected from the PRI to mount an opposition presidential bid, the anti-pri resentment 135 Interview with Gómez Tagle. 221

230 within civil society and the broader electorate found a positive outlet. A heterogeneous front of left elites and civil society actors, despite weak horizontal linkages and manifold differences, united around a single cause: to defeat a dominant party that, in their view, had committed to a failed, right-wing economic model and proven unresponsive to internal and external demands for economic reform and social justice. Cárdenas offered the national clout necessary to make this cause reality. Paradoxically, the PRI s fraudulent, repressive response to cardenismo strengthened the left opposition by hardening activist commitments. During Salinas s sexenio, national PRI politicians and leaders and subnational authorities stole the presidency, stole governorships, stole mayoralties, killed scores of left activists, manipulated the media establishment, and channeled vast resources, all in order to cow and defeat the FDN and PRD. For perredistas, the Salinas administration and subnational PRI oligarchies, during this period, went well beyond their original crimes, revealing that they would stop at nothing to retain power. As the political establishment employed the full powers of the Mexican state to keep the left in the political wilderness, activist anger and defiance mounted, and mobilization increased. Fraud across levels of government provoked a steady stream of left mobilization from 1988 to the mid-1990s. As discussed earlier, Cárdenas s defeat provoked a sustained mass movement that eventuated in the creation of the PRD. Also in 1988, Andrés Manuel López Obrador, the FDN s gubernatorial candidate in Tabasco, officially lost to the PRI s Salvador Neme Castillo, alleged fraud, and organized a series of mass actions to protest the new government, including rallies and blockades of highways and gas stations. Throughout 1989 and 1990, as the PRI systematically defrauded the PRD in municipal elections across the country, PRD supporters and activists mobilized at the local level, staged marches, blocked highways, and occupied city halls. 136 In 1991, AMLO organized his first exodus for democracy from 136 Media broadcast these actions in order to reinforce perceptions of the PRD as destabilizing. 222

231 Tabasco to Mexico City in protest of municipal-level fraud. The same year, in San Luis Potosí, PRD activists carried out a long series of mass demonstrations and marches to protest the illegitimate triumph of Fausto Zapata (PRI) over the PRD s Salvador Nava. 137 Similar demonstrations occurred after gubernatorial elections in Michoacán in 1992 and Guerrero in In 1994, AMLO lost his second bid for the Tabasco governorship, this time to Roberto Madrazo (PRI). In 1995, presenting copious evidence of electoral fraud and other abuses, AMLO organized his second exodus for democracy, leading 2500 PRD activists 138 to Mexico City to protest the Tabasco results as well as systematic fraud at the national level. 139 To an even greater extent than systematic fraud, however, the use of violence and repression by subnational PRI authorities strengthened the PRD cause. Violence against the left galvanized the PRD base, motivated new entrants, 140 and generated cohesion. According to the testimony of numerous party founders, the repression of the Salinas years hardened PRD activists. Instead of retreating or exiting politics, most perredistas became more resolute in the face of violence. 141 According to Borjas (2003), the repression exercised by the government...contributed to unifying and strengthening the convergence of distinct groups and citizens generally who resolved to unite to resist governmental violence (341). In 1992, Porfirio Muñoz Ledo wrote in Proceso: Every time the PRI confronts us directly...the party 137 Tellingly, the PRI refused to cede the governorship to the PRD after Fausto Zapata s resignation, even though in the same year, the PRI did cede the governorship of Guanajuato to PAN candidate, Carlos Medina Plascencia, in response to weaker anti-fraud protests by PAN activists. 138 Borjas (2003, vol. II): The protesters also addressed issues such as the potential privatization of Pemex. 140 In addition to galvanizing the existing base, the PRD s David and Goliath-like struggle against the PRI may also have inspired and activated previously inactive individuals and supporters. Borjas (2003, vol. II) argues that the PRD s elevated territorial coverage in Michoacán, Guerrero, and Chiapas can be attributed precisely to repression as a source of party members cohesion, resistance, and activism (vol. II: 245). 141 Interview with Borjas. 223

232 unifies, strengthens inside, works toward a tighter coordination and deeper solidarity. It isn t new that adversity unifies. 142 For many perredistas, PRI violence transformed the conflict between the political establishment and left opposition into a war. When asked why the PRD did not fold after the disastrous 1991 elections (about which more below), one party founder from San Luis Potosí replied simply: We were at war. 143 In a 2010 interview, party founder and early federal deputy Carlos Navarrete stated: [The Salinas sexenio] was a very hard time because our struggle was against the current. They were times of persecution, of hundreds of dead activists. They were times in which they stole elections from us, covering the width and depth of the country. They were times in which the government besieged us. They were times of resistance, fundamentally, of not giving up, of maintaining and raising our flags. 144 Martínez (2005) observes that, as late as 1994, most of the PRD opposed negotiation with the PRI due to the party s warlike relationship (beligerante trato) with Carlos Salinas (70-1). In major speeches, party leaders frequently paid tribute to fallen perredistas. In his acceptance of the PRD s nomination for the 1994 presidential campaign, for example, Cárdenas commemorated the 239 party members murdered thus far (Borjas 2003: 512): We must honor those who, with their sacrifices and their lives, made it possible for us to arrive at this day: hundreds of deaths for democracy across the country have been a harsh price paid by all Mexican people. 145 In 2010, the PRD produced a hardback commemorating the party s twenty-year anniversary. There were two dedications, the first to those who lost their lives in the struggle to build the Party of the Democratic Revolution (González et al., eds., 2010) Proceso no. 820, 7/20/1992, p Interview with Nava. 144 Interview with Navarrete in González et al., eds. (2010: 265). 145 Proceso, no. 1101, December 7, 1994, pp Cited in Borjas (2003, vol. II): The second was to the militancia. 224

233 The fraud and repression of the Salinas years remain salient for the PRD s early elites and activists and continue to inform their perspectives, behaviors, and collective identity. Party founder and current party president Jesús Zambrano ( ) recalled in a 2010 interview that following the imposition of Salinas, neoliberalism became consolidated, and the left experienced the most brutal repression in its history, with over 600 deaths during the Salinas sexenio. Party members González et al., eds. (2010) describe the Salinas years as the hardest moments for the PRD. In this period we have 500 PRD deaths, the stolen election in Michoacán, such a grave problem in the Guerrero election, and the exodus for democracy in Tabasco. 147 Many perredistas of long standing still consider electoral fraud endemic in Mexico. 148 Others lament that newer generations of perredistas lack an awareness of the party s adverse origins. In the words of party founder and early Texcoco mayor, Horacio Duarte: When you talk to [19- and 20-year olds] about how hard it was to build the PRD opposite a PRI government, they don t understand you because they didn t live through that period. 149 * * * In summary, due to circumstances of adversity and the elite incentives, selection pressures, and activist orientations that flowed from these circumstances, the early PRD constructed a strong organization composed of committed activists. By the mid-1990s, the party had recruited over one-million members and built municipal offices in most Mexican municipalities. Early perredistas were primarily left ideologues and principled opponents of the PRI. They were not driven by selective incentives, as PRD elites had few selective incentives to offer. 147 González et al., eds. (2010: 66). 148 After the 2006 presidential election, perredistas, despite a high degree of skepticism within the broader electorate, overwhelmingly believed that Andrés Manuel López Obrador had lost due to fraud. For two months after the election, hundreds of thousands of perredistas, perhaps millions, participated in mass demonstrations, rallies, marches, and encampments. 149 Interview with Duarte in González et al., eds. (2010: 249). 225

234 In contrast to patronage-seekers, the local leaders and activists who built the PRD were not likely to defect from the partisan left in the absence of short-term payoffs or government spoils. They genuinely believed in the party s cause: to defeat the violently authoritarian PRI, halt the advance of neoliberalism, and usher in an era of progressive, democratic change in Mexico, all under the moral leadership of Cuauhtémoc Cárdenas. Electoral crisis and survival in 1991 and 1994 Greene (2007) argues that the PRD s early joiners hurt the party by supporting unelectable candidates and promoting radical, unpopular policy stances. This account passes over an important point: the commitment of the PRD s early joiners, rooted in opposition to PRI authoritarianism and neoliberalism and hardened by PRI hostility and violence, made the PRD more durable during the formative phase. Early activists did not join the PRD because they expected a quick, easy route to national power and the associated spoils. They believed in the PRD s cause and accepted, even welcomed, the prospect of an extended, hard-fought struggle. Partisan belief and commitment extended their time horizons, steeling them against early electoral setbacks. In particular, organizational strength and activist commitment contributed decisively to the PRD s survival of early electoral crises in 1991 and The PRD suffered major electoral setbacks in both years, first in the Chamber of Deputies and second in the presidential election. In both cases, networks of committed activists rebounded and redoubled their efforts in the party s territorial strongholds. The 1991 Chamber of Deputies election Memories of the 1988 general election lifted the early PRD s electoral expectations. The FDN elected almost thirty percent of Mexico s federal deputies, and Cárdenas won the pres- 226

235 idential contest on most accounts. Subsequently, perredista candidates registered strong performances in the party s first elections at the subnational level. Despite a slew of disadvantages, the PRD won scores of mayoralties in 1989 and Absent fraud, the party would have won many additional mayoralties and might have won several governorship (e.g., Tabasco, San Luis Potosí). In view of these encouraging early results, PRD leaders and activists looked to the 1991 congressional elections with optimism. Despite a highly uneven electoral playing field, leaders and activists shared...a certainty that the party would win broad representation in the Mexican Congress (Borjas 2003: 388). A strong performance in the election, they believed, would enable the PRD legislative bloc to stymie the Salinas agenda, pass institutional reforms leveling the electoral playing field, and thus lay the groundwork for a Cárdenas victory in This optimism represented the...point of convergence within the party organization (Borjas 2003: 288). The PRD performed abysmally. The party lost seats 150 and finished a distant third, with a mere eight percent of the vote. The PAN maintained a vote share of eighteen percent, while the PRI jumped from fifty-one to sixty-one percent. 151 Taken together, the constituent party s of the former FDN increased their vote share and defeated the PRD. The PRD also failed to elect a governor for the third consecutive year. The PRI won all seven 1991 gubernatorial elections (Campeche, Colima, Guanajuato, Nuevo León, Querétaro, San Luis Potosí, and Sonora), although it ceded the governorship of Guanajuato to the PAN in response to anti-fraud protests. The 1991 results dashed perredistas expectations and threatened the party s survival. Prominent PRD leaders from the social left, citing PRI abuses, manipulations, and violence during 150 The PRD had entered the 1991 midterms with fifty-five deputies 151 In the election, the party performed well above the party average in its organizational redoubts. Specifically, federal deputy candidates were most successful in Michoacán, Guerrero, and, to a lesser extent, Oaxaca, Tabasco, the Federal District, Mexico state, Hidalgo, Tlaxcala, and Nayarit. 227

236 the campaigns, advocated that the PRD completely withdraw from the electoral sphere. In an early 2000s interview, Cárdenas described 1991 as the most challenging episode in the PRD s history: The objective of the 91 midterms was the real, effective liquidation of the PRD... [T]his was the PRD s hardest electoral moment. 152 Less than a month after the election, Porfirio Muñoz Ledo publicly stated that the PRD s continued electoral significance might hinge on an alliance with the conservative PAN an unthinkable prospect for much of the PRD rank-and-file. 153 Bruhn (1998) writes that the PRD s dismal performance [in 1991]... damaged its credibility as a threat to the PRI...and led many to question its very survival (250). Borjas (2003) states that after the election, it was completely undeniable that the survival of the PRD...was in danger (457-8). 154 After the election, party leaders, despite their concerns about the party s viability, exhorted the base to regroup. Leaders urged militantes, in particular, to redirect their attention toward upcoming subnational contests, municipal and state. In its internal diagnosis, the National Executive Council emphasized that the PRD continued to suffer from great shortcomings in the organizational domain and resolved, moving forward, to prioritize territorial expansion from the regional to precinct level, and to identify areas of infrastructural weakness with X-ray precision. 155 Cardenas himself proposed a major organizational effort, and Muñoz Ledo concurred, proposing that the party expand its territorial reach, in electoral and organizational terms, by forging broader alliances. 156 Independently of top-down exhortations, however, the 1991 crisis actually galvanized much of 152 Interview with Cuauh emoc Cárdenas in Borjas (2003, vol. II): Proceso, no. 776, September 16, 1991, pp The party, Borjas (2003) observes, suffered a severe setback (retroceso) not only in relation to the seats obtained by the constituent parties of the FDN, but even relative to itself, given that it did not manage to hold the fifty-five seats obtained by the FDN members who had become perredistas (408). 155 Proceso, no. 776, September 16, Proceso no. 773, 8/26/1991: 10-11, Proceso no. 776, 9/6/1991:

237 the PRD base. As described by the editors of the PRD s twenty-year commemorative: 1991 represented for the PRD its first political setback but at the same time the consolidation of an iron-willed base (militancia férrea) (González et al., eds., 2010: 66). Early perredistas viewed the PRI s overwhelming victory in 1991 as illegitimate: the product of repression, massive state spending, a corrupt media establishment, and fraud. By and large, these activists opted to continue the struggle. One party founder and activist, recalling the mentality at the base level after the 1991 election, emphasized that the PRD was at war. The PRI had murdered scores of activists and, just as in 1988, defrauded the national electorate. The PRD had not actually lost. For these reasons, the interviewee stated, the elections did not depress early activists ánimo (spirits). 157 Another founder and early federal deputy with social left origins expressed a similar sentiment. The elections, he stated, were not a huge disappointment because there had not been real competition. Priístas had used every conceivable lever of state to win. They had cheated and used the state apparatus and media. Local chiefs with impunity had killed activists just for defending the vote. PRI hostility motivated the militancia. PRD activists did not even entertain the thought of giving up after They had to get organized, as there was no other way to defeat the PRI. 158 The PRD survived the electoral crisis of 1991 due to the persistence of activist networks in the party s territorial redoubts. In the gubernatorial elections of 1992 and 1993, the PRD s first major post-1991 contests, the PRD, despite failing to win a single governorship, registered strong performance in numerous states. In particular, PRD candidates performed well above the party average in organizational strongholds such as Michoacán, Guerrero, Nayarit, Veracruz, and Zacatecas (Borjas 2003: 490-1). The PRD also advanced at the 157 Interview with Nava. 158 Interview with Saucedo. 229

238 state-legislative level in most territorial strongholds. In 1992 and 1993, the PRD s top leadership gave public voice to the confrontational, defiant attitude that predominated among local party activists. In mid-1992, Muñoz Ledo stated in Proceso: What irritates Salinas is the PRD s position... the attitude of democratic intransigence... It bothers him that we re incorruptible. That s what pains him the most. What most irritates him is that we can t be bought, and don t even feel afraid, to such a degree that we go looking for fear recklessly. 159 Along similar lines, Cárdenas stated in 1993 that Salinas must be annoyed...because he has not been able, with all the state s resources, to crush the PRD or its leaders. 160 The 1994 presidential election The PRD survived the 1991 congressional election due to organizational strength and activist commitment. Local networks of base-level perredistas regrouped and continued their party work, expanding the grassroots organization and campaigning for subnational PRD candidates. In the 1994 presidential election, however, the PRD suffered another electoral crisis, testing the PRD s durability a second time. In a sense, Cárdenas s landslide loss in 1994 posed a more significant challenge to the PRD than the 1991 electoral crisis. The prospect of a Cárdenas victory and, with it, the completion of Mexico s transition to democracy and an historic victory for the Mexican left had animated perredistas since the party s founding. CEN member Jesús Ortega wrote in Proceso in August of 1995 that the PRD had possessed only one objective until that point: to elect Cárdenas president the previous year. 161 According to Martínez (2005), the PRD was conceived around one person and one idea: to carry Cuauhtémoc Cárdenas to 159 Proceso no. 820, 8/30/ Proceso no. 820, 8/30/ Proceso, no. 980, August 14, 1995, pp Cited in Borjas (2003):

239 the presidency (97). Borjas (2003) writes that rescuing, or recovering, the presidency for Cárdenas constituted the PRD s central short-term objective as of 1993 and, more generally, one of the main ideas...guid[ing] the actions of the PRD in its initial years (456, 507). Rodríguez (n.d.) describes the goal to win the 1994 presidential election to take back what they stole from us in 1988 (257) as the early PRD s central reason for existence. 162 Throughout the early 1990s, numerous perredistas did not believe that the PRD would survive if Cárdenas lost in Further, according to pre-election conventional wisdom within the PRD, Cárdenas would prevail if the party s ground forces, unlike in 1988, did a thorough job of defending the vote. 164 The PRD s 2010 hardback commemorative recalls the atmosphere at the II National Congress in mid-1993: The spirit of the second congress was different from It was of resistance, of electoral strength reduced to eight percent, but with a militancia convinced that it would win Borjas (2003) writes that PRD leaders and militantes were sure that in 1994, circumstances would be different (507). But Cárdenas lost in a landslide. He placed a distant third with seventeen percent of the vote. 166 That Cárdenas not only lost, but received under one-fifth of the valid votes, damaged the party s credibility and self-conception as a serious challenger to the PRI. Party moderates argued that the PRD lost, fundamentally, because the Cárdenas campaign had espoused radical views, which alienated moderate left voters, and relied on antiquated campaign tactics such as rallies, which limited Cárdenas s exposure (Greene 2007). Porfirio Muñoz Ledo, for 162 Personal communication with Rodríguez, August 1, Personal communication with Bruhn, June 10, Personal communication with Rodríguez, Aug. 1, González et al., eds. (2010: 66). 166 In the concurrent Chamber of Deputies election, the PRD also placed a distant third. Although the PRD gained seats in the Chamber of Deputies, party leaders had anticipated a stronger result due to presidential coattail effects. 231

240 example, criticized the Cárdenas campaign s programmatic decisions, especially to support the Zapatista Army of National Liberation (EZLN) and oppose the North American Free Trade Agreement (NAFTA). 167 Cárdenas himself highlighted the campaign s overreliance on rallies (mítines) and underutilization of more modern, efficient forms of campaigning such as mass media appeals. 168 To be sure, Cárdenas s loss deflated hopes and led to anger, disillusionment, and recriminations. As in 1991, however, the PRD militancia more radical, by and large, than the party moderates who criticized the Cárdenas campaign remained unfazed. Local PRD activists viewed Cárdenas s crushing defeat as part of a long pattern of illegitimate electoral outcomes, based on media ostracism and hostility, massive resource asymmetries, activist repression, and (rightly or wrongly) fraud. 169 The PRI s hostility and perceived abuse of power thus remained a galvanizing force. During a discussion of the 1994 election, one party founder stated that the playing field remained highly uneven, and the PRD remained at war. Perredistas, he stated, owed it to their fallen comrades to soldier on and eventually defeat the PRI. 170 As had occurred in 1991, PRD activist networks in the party s territorial strongholds, following Cárdenas s 1994 defeat, redirected their attention and energies toward subnational elections, both for electoral mobilization and (in Tabasco) post-electoral civil resistance. As noted earlier, the year 1995 gave rise to the second and more heavily publicized exodus for democracy, in which AMLO led a march from Tabasco to Mexico City in protest of both 167 Muñoz Ledo also wrote that the PRD had underestimated the PRI s capacity to manipulate the electoral process. Proceso, no. 964, April 24, 1995, p. 23, and no. 931, September 5, 1994, pp Cited by Borjas (2003): Proceso, no. 933, September 19, 1994, p. 10, and no. 940, November 7, 1994, p. 39. Cited by Borjas (2003): Interviews with Nava and Saucedo. 170 Interview with Nava. 232

241 state and national fraud. In territorial strongholds, the PRD performed well above the party average in 1995 and 1996 state and local elections, including the 1995 gubernatorial race in Michoacán, the 1995 and 1996 state legislative contests in Oaxaca and Guerrero, and the 1995 and 1996 municipal elections Chiapas, Michoacán, Oaxaca, Mexico state, Guerrero, Hidalgo, and Nayarit. 171 Thus, as with the electoral setback in 1991, after the 1994 presidential election, PRD activist networks in the party s territorial redoubts regrouped and shifted focus to subnational contests. Asked to explain the PRD s resilience after the 1994 election, a party scholar based in Mexico stressed the importance of the PRD militancia. Loyal networks of activists, he stated, never accounted for most PRD votes, but by remaining committed and active through disappointing times, they gave the PRD a sturdy electoral floor and enabled the party to overcome crises like Cárdenas s 1994 loss. 172 PRD organization-building during the late 1990s During the second half of the 1990s, the PRD began to engage aggressively in top-down and more mercenary forms of organization-building. This initiative formed part of a broader strategy to improve the PRD s future performance in major elections. Under the party presidency of AMLO (1996-9), the National Executive Committee (CEN) made a series of tactical compromises to position the PRD for victory in the 1997 Federal District mayoral election, a slew of late 1990s gubernatorial elections, and the 2000 presidential election. Controversial among PRD purists, these compromises included major new mass media initiatives and, in the domain of organizational expansion, the large-scale use of paid activists and the frequent co-optation of external candidates and activist networks. 171 Borjas (2003, vol. II): 94, Interview with Ortega. 233

242 AMLO s media strategy arose in part from the 1996 political reforms and the opening of private media in the late 1990s, but also from the recognition that ground organization and rallies alone did not deliver major executive victories. Cárdenas s 1994 defeat raised awareness within the PRD leadership that, at least in major national elections, traditional party organization and campaign events could not substitute for media. Citing the impact of regime capacity and media attacks, one party leader stated in Proceso: We thought that just with rallies we were overcoming the actions of the [government] apparatus in the electoral process. 173 As party president, AMLO impressed upon the PRD that success in the largest elections would require a greater willingness and ability to use mass media. His internal administration thus poured campaign resources into media advertisements. Simultaneously, however, the PRD retained its organizational ethos and sought to continue its organizational expansion even as it embraced media. In particular, AMLO sought to beef up the PRD s electoral infrastructure in historically problematic localities and regions, where base-level party networks still had not sprouted up. By the mid-1990s, the PRD claimed over one million members and formal branches in most of the nation s municipalities, but this left much of the country untouched. In most Mexican states, the PRD lacked professional cuadros capable of governing at the state level or ground organizations large enough to drive successful gubernatorial campaigns. Aided by the influx of public funding in 1997, AMLO s CEN financed the creation and development of the Brigadas del Sol (Sun Brigades), a large army of new militantes activated for electoral campaigns and remunerated for their services, in contrast to the PRD s initial activist cohort. The PRD allocated thirty percent of the 1997 campaign fund to the Brigadas. By the end of the year, the Brigadas included 63,000 paid activists, one for every electoral precinct in Mexico (Borjas 2003, vol. II: 124-5). In 1998, territorial expansion remained a top party priority. Cárdenas stressed at the party s 173 Proceso no. 930, 8/29/

243 fourth National Congress in March of that year: The road that the PRD must travel in the coming months and years is not easy. The party must keep building, developing, and consolidating itself across the country; creating new base-level and municipal committees. 174 In order to increase the PRD s territorial implantation in organizational and electoral terms, AMLO initiated the practice, in states with a weak PRD presence, of incorporating external candidates, often defecting priístas, with their own local and regional networks and machines. In 1998, gubernatorial elections took place in nine states: four in the North (Chihuahua, Durango, Sinaloa, Tamaulipas), two in the Central region (Aguascalientes, Zacatecas), two in the East (Puebla, Tlaxcala), and one in the South (Oaxaca). In eight of the nine states, the PRD fielded external candidates. Five of the PRD s external gubernatorial candidates defected from the PRI to run on PRD tickets. AMLO s electoralist tactics provoked heated internal criticism, nationally and subnationally, from PRD leaders and activists of radical and social left origin. These leaders rejected the use of selective incentives to fuel activism and the co-optation of external candidates, especially ex-priístas with little or no connection to Cárdenas s original Corriente Democrática. Yet the tactics paid dividends, both organizationally and electorally. It was under AMLO s party presidency that the PRD won its first major executive victories: the 1997 DF mayoralty and three governorships in All three successful gubernatorial candidates were external and ex-pri: Ricardo Monreal in Zacatecas, Alfonso Sánchez in Tlaxcala, and Leonel Cota in Baja California del Sur. In Zacatecas, the PRD, prior to inheriting Monreal s regional machine, claimed offices in only twenty-one of the state s fifty-seven municipalities. Monreal brought an extensive network of loyalties with him, helping bolster the state PRD organization (Borjas 2003, vol. II: 211). As of 1998, the PRD claimed two-million members, six-thousand comités de base, full municipal office penetration percent in the Federal District 174 Citation needed. 235

244 and six Mexican states, over 85-percent penetration in thirteen states, and nearly 64-percent penetration nationally Borjas (2003 II): 65, 74, 244-5,

245 Chapter 5 Avoiding schism in the early PRD: the indispensability of Cuauhtémoc Cárdenas Adverse circumstances led to the emergence of a strong PRD organization composed of committed activists. While critical for the PRD s survival of early electoral crisis, organizational strength and activist commitment also brought risks of fragmentation and schism, for two reasons. First, as suggested earlier, the early PRD s committed activists and leaders, for the most part, were ideologues, not pragmatists (Greene 2007). Although ideology extends individuals time horizons and thus strengthens member loyalty amid crisis, ideologues are also more likely than pragmatists to value ideological and programmatic purity over electoral success. Many ideological parties and political movements even relatively narrow, homogeneous ones suffer from low cohesion and bitter divisions (e.g., Latin American communist parties). In the early PRD, the strong presence of ideologues, other things equal, militated against compromise and contributed to sectarianism and factional conflict. Second, the PRD was far from homogeneous. It is almost never feasible to build a strong party organization without incorporating a highly heterogeneous set of individuals and organizations. The PRD, in order to establish party nuclei and branches in large swathes of the national territory, had to allow for considerable internal difference, and not just in ideological terms. In a single umbrella party, or partido-escoba, 176 the early PRD brought together social movements, civil society organizations, regional PRI structures, traditional Marxist structures, and several small political parties. Although these groups shared the objective of electing Cárdenas and, in the process, democratizing Mexico and halting the country s neoliberal turn, they differed in their long-term ideological visions, short-term programmatic and tactical preferences and priorities, class makeups, regional profiles, organizational subcultures, and personal loyalties. 176 Borjas (2003) uses this term (258). 237

246 Within the traditional Marxist left alone, one of just three central PRD feeder categories, all of the above differences existed. The constituent parties of the PMS, overwhelmingly the PRD s main source of traditional Marxist elites, cuadros, and organizational structures, spanned the full range of Marxist currents, possessed their own partisan cultures, and professed loyalty to competing leaders. The PMS succeeded the PSUM, itself an almost intractably heterogeneous Marxist coalition. The declaration of principles of the PSUM had reflected an almost impossible amalgamation of ideologies. In a single party, the communist, socialist, lombardista, 177 Trotskyist, liberal, and revolutionary nationalist [lefts] had come together (Hérnandez 2010: 34). In addition to absorbing the PSUM, the PMS, as discussed earlier, incorporated six separate Marxist parties and organizations, some firmly institutionalist and reformist (e.g., Jesús Ortega s defecting PST faction), others less so (e.g., PPR, MRP). 178 During its short existence, the PMS lacked a common identity to erase the loyalties of members of its component parties (Bruhn 1998: 163). Thus, the traditional Marxist left, just one of the PRD s three major feeder categories, lacked cohesion. Moreover, the central point of convergence within this PRD feeder category established participation in democratic institutional politics set it apart from a second major PRD feeder category: the extraparliamentary social left. Gradualist rather than revolutionary, the constituent parties of the PMS, for decades, had embraced electoral competition as the primary means for advancing left ideals and interests. In contrast, the PRD s feeder movements, organizations and parties from the social left, prior to 1988, had either disdained democratic politics or simply not participated. These actors embraced tactics of mobilization and, for the most part, categorically rejected negotiation and dialogue with the PRI. 177 After Mexican labor leader Vicente Lombardo Toledano ( ). 178 The entry of groups such as the PPR and MRP marked a hitherto rare fusion of social and traditional left forces. Predictably, leaders of the PPR and MRP revived an argument that the [PMS] had supposedly overcome: the commitment to legality (Martínez 2005: 92). See also Rodríguez (n.d.):

247 Despite sharing an extraparliamentary orientation, groups on the social left differed widely along dimensions of region, class, and ideology. As noted earlier, they included ex-guerrilla nuclei, rural unions, popular urban movements, teachers unions, blue-collar unions, student associations, NGOs, and post-materialist organizations. Some were anarchist, others nationalist. Most were Marxist but, within Marxism, differed in sectarian orientation, ranging from Trotskyist, Maoist, and Guevarist to revolutionary Christian (Martínez 2005: 53). The third major feeder category of the PRD, the Corriente Democrática (CD), contained ideological divisions at the elite level and distinct leadership loyalties at the base level. At the elite level, CD leaders divided into two camps: those centrally concerned with mass mobilization and confrontation (e.g., Cuauhtémoc Cárdenas), and those focused on reform through institutional politics (e.g., Porfirio Muñoz Ledo). This difference mirrored that between the confrontational, movimientista social left and the predominantly reformist, institutionalist PMS. At the base level, the PRD s two strongest ex-pri feeder networks, in Michoacán and Tabasco, supported different leaders, Cuauhtémoc Cárdenas and Andrés Manuel López Obrador. For the individuals within these organized bases, loyalty to the relevant leader (Cárdenas, AMLO) trumped commitment to the broader array of forces in the PRD. Prior to the PRD s creation, horizontal linkages, both within and across PRD feeder categories, were often superficial or nonexistent. Where relations ran deep, they were typically characterized by high levels of conflict and low levels of routinization. Ties between defecting PRI networks, for example, barely existed. Regional ex-pri networks in Michoacán, Tabasco, and Mexico City, prior to the PRD s creation, shared a party label but had existed independently, with little interaction, for decades. The sprawling social left, having eschewed elections during the 1970s and 1980s, never had an incentive to unify at the national level before the PRD s founding. Thus, social left movements and organizations possessed little or no experience of negotiating, bargaining, and resolving conflicts with each other under 239

248 the umbrella of large national coalition movements and organizations. In particular regions, social left groups often did possess ties that predated the FDN and PRD. In most cases, however, these ties had given way to bitter sectarian division by the PRD s founding. The social left in the state of Guerrero, a major bastion of the early PRD, provides a case in point. In the 1960s and 1970s, Guerrero constituted the primary battleground of the dirty war between Mexico s rural guerrilla insurgency and the state, and during this time, numerous guerrilla groups and sympathetic left organizations pooled resources and joined forces. After the guerrillas decisive defeat in the 1970s, however, alliances frayed, and Guerrero s social left fragmented into a multitude of conflicting sects with their own leaders, strategies, and local and sectoral strongholds (Fernández 1993). Although the traditional Marxist left did, through the creation of the PSUM and PMS, make a modicum of progress toward unification before the PRD s founding, horizontal relations within the traditional Marxist left remained conflictual and weakly routinized at the base and elite levels. Excluded from the unification process, activists and members in the PMS s feeder parties did not have time to redefine their models of activism and organization in a democratic manner (Rodríguez 2010: 258). At the elite level, informal negotiations rather than formal procedures determined the allocation of candidacies and internal positions (Bruhn 1998: 163). Until the PRD s creation, the PMS remained a hostage of its internal divisions. Heterogeneity...blocked cohesion (Martínez 2005: 92). Prud homme (1997) summarizes: As a party, the PMS never managed to become institutionalized: the most disciplined cuadros of the PSUM and PMT coexisted with a multiplicity of leaders from the small, radical left organizations (12). Finally, linkages across the PRD s three broad feeder categories were also weak or nonexistent. Historically separated by distance and ideas, the CD-aligned dissident PRI, traditional Marxist left, and social left, before Cárdenas s presidential bid, had not joined in collec- 240

249 tive movements or organizations. The year 1988 marked the first major confluence of these categories. In sum, the early PRD, in absorbing most of Mexico s traditional Marxist left, most of its social left, and multiple regional PRI structures, inherited a geographically dispersed and highly heterogeneous set of movements, organizations, and partisan networks characterized by superficial or nonexistent horizontal relations or, where relations ran deeper, high levels of sectarianism and conflict and weakly routinized collective decision-making procedures. Of necessity, the PRD, on creation, inherited all of these characteristics. Barriers to unity in the early PRD Much like Brazil s PT and Peru s IU, the PRD, from its founding statutes, was a party of factions, not with factions. Founding statutes structured the PRD organization as a presidential system mediated by corrientes, formally recognized factions often described as the PRD s parties or, in informal parlance, tribus (tribes). Throughout the PRD s formative years, in order to receive candidacies or ascend in the national or subnational party apparatus, one had to belong to or ally with a corriente. In large measure, corrientes received primacy in the early PRD because of the party s PRI origins, and because wellorganized groups within the PRD, especially from the PMS and CD, wished to place a check on Cárdenas s internal power (Martínez 2005: 101). Except during national Congresses, power within the PRD resided in the National Executive Committee (CEN) and the larger National Council (CN). The president of the party was the chief executive and head of the CEN. The CEN consisted of numerous secretarías, the equivalent of PRD ministries. The formal structure of state and municipal party organs paralleled that of the national organization. Power at these levels resided in state and 241

250 municipal executive committees and councils. National Congresses, convoked as necessary, were the PRD s supreme authority and consisted of delegates selected, through varying mechanisms, by subnational organs. Given the level of heterogeneity within each feeder category of the PRD, the early PRD organization was extremely heterogeneous, with many leaders, partisan and organizational cultures, and networks of personal loyalty coexist[ing] (Prud homme 1997: 12). Martínez (2005) describes the early PRD as an explosive cocktail that mixe[d] communists and Trotskyists, ex-priístas and ex-guerrillas, nationalists and internationalists (59). 179 In broad terms, the internal functioning of the PRD s early municipal organs, state organs, and national secretarías (PRD ministries) varied depending on which leader/s, backed by which PRD feeder organization/s, controlled them. The result was multi-layered organizational heterogeneity, from municipal organ to municipal organ, from state organ to state organ, and from national secretaría to national secretaría. Reflecting a precarious union, 180 these organs and agencies possessed weak or nonexistent horizontal linkages, a fact emphasized in Borjas (2003) and Rodríguez (n.d.). Within individual party organs, especially the national and top subnational executive committees and councils, conflict and division ran rampant between leaders and between corrientes. Global ideological conditions fixed the boundaries of this internal conflict and division. The PRD was founded the same year that the Berlin Wall fell. The collapse of the Soviet Union undermined Marxism as a global ideology and, consequently, shaped the ideological struggle within the PRD. Debates did not center on grand questions of capitalism versus communism. Leaders and corrientes did differ, however, on fundamental questions of program, 179 The diversity of the groups and organizations that supported Cárdenas (and the left in general), writes Bruhn (1998), made profound internal divisions very likely (165). 180 Borjas (2003) writes that due to the diversity of PRD feeder groups, high heterogeneity and a precarious union obtained between secretarías, between state committees, and between municipal committees (450; see also 301, 518). 242

251 internal procedure, and above all, the proper tactics and timeline for democratizing Mexico and rolling back neoliberalism. 181 At the national and subnational levels, leaders competed for coveted candidacies and control of the internal apparatus, both for personal gain and to advance factional agendas. From these struggles, clear winners and losers emerged. Early state PRD organs, for example, often became terrains of bitter sectarian conflict between dizzying arrays of left forces. In Guerrero, for example, seven distinct early PRD projects emerged, dominated by moderate leaders in some cases (Zenón Santibáñez, Eloy Cisneros) and radical leaders in others (e.g., Félix Salgado, Rosalío Wences), and characterized in every case by different combinations of social and traditional left actors, 182 including ex-guerrilla groups, peasant unions, teachers unions, and university networks linked to the powerful 183 Autonomous University of Guerrero. 184 Although these PRD projects operated under the PRD label, the party label hid the competing strategic projects that had pulverized the left in the 1970s and 1980s, as well as all the accumulated resentments (rencores) of that period (Fernández 1993). Similar internal divisions obtained at the state level in the PRD s other early bastions, including Chiapas, Mexico state, 185 Michoacán, Tabasco, 181 Javier Hidalgo emphasized this point in an interview with the author. 182 Unlike in Michoacán and Tabasco, in Guerrero, defecting PRI networks played a very marginal role in building the PRD (Fernández 1993). 183 Fernández (1993) observes that, given the weakness of Guerrero s economic sector, Catholic Church, national political class, and party system, the UAG played a major role in mediating relations between the state government and population. 184 The seven PRD projects identified by Fernández (1993) had the following leaders: Orthón Salazar, teachers union leader and ex-pcm member, in the Montaña region; Félix Salgado, radical, former PFCRN member, and early FDN/PRD deputy and senator, in the North region; Zenón Santibáñez, moderate dialoguista, former PFCRN member, and bitter rival of Félix Salgado, in the Tierra Caliente region; Eloy Cisneros, ex-guerrilla with ties to the Autonomous University of Guerrero, in the Costa Chica region; different alliances of ex-guerrilla and universitarios in the Costa Grande and Acapulco regions; and Rosalío Wences, former PCM member, rector of the strategically important Autonomous University of Guerrero ( ), and arguably Guerrero s most important PRD leader in the early 1990s, in the Central region. Wences chose not to run for governor of Guerrero in 1994, preferring to run for rector a fourth time (Fernández 1993). 185 In Mexico state, the social left, traditional Marxist left, and former PRI converged in the FDN and early PRD, and local factions tended to mirror political divisions among the old left parties and leaders [and] 243

252 Tlaxcala, and Zacatecas. 186 At the national level, however, the PRD s internal struggles received their fullest expression. The national leadership and associated corrientes represented all of the PRD s major points of view and all three of its feeder categories (the former PRI, the traditional Marxist left, and the social left), along with each category s internal contradictions. Moreover, internal conflict resolution and decision-making were weakly routinized. Perhaps no single fact illustrates this more clearly than the frequency of national Congresses. In order to change core party statutes and address various forms of internal dysfunction, the PRD convoked six congresses in the first twelve years of its existence. Brazil s PT, in contrast, held only one Congress in its first twelve years. Subnational organs typically followed their own rules. In Borjas s (2003) words, state and municipal committees functioned according to who led them and not based on established norms and procedures (519). Decision-making practices and outcomes in the critical areas of programmatic development, candidate selection, delegate selection, the distribution of internal positions, membership criteria, and activist formation varied from one subnational organ to the next. Where national party statutes dictated to subnational organs in these areas, subnational organs often flouted the relevant statutes, almost always without consequence. The national party also exercised little control over subnational politicians and national legislators, who rarely received serious sanctions if they deviated from the formal or informal party line. More importantly, within the National Executive Committee and Council, the PRD, during its formative years, never consolidated formal procedures for resolving the most critical ex-priístas. Prevailing divisions reflected, remarkably, the original political experience of an individual - ex-communists versus ex-pmt members versus MASistas, and so forth (Bruhn 1998: 191). 186 For analyses of the PRD s development in most of these states, see the chapters in López and Cadena (eds., 2013). 244

253 conflicts between leaders and corrientes. As of the mid-1990s, the national leadership had still failed to reach agreement on a common national party program. 187 Throughout the 1990s, the PRD did not have stable formal procedures for selecting candidates, for selecting its own president, or for distributing positions within the party apparatus. Typically, the national party leadership depended on informal negotiations to distribute candidacies and internal posts. When the leadership adhered to formal procedures, elites often determined the specific procedure through informal negotiation and on a case-by-case basis. 188 National leaders regularly flouted the formal rules of internal competition, such as the ban on running in consecutive PRD presidential elections. Moreover, procedures intended for long-term use rarely endured more than a few years. The party s procedure for electing its own president, for example, changed over the course of the 1990s, beginning as a congressional vote and shifting to a closed primary. 189 Procedures for determining the composition of the CEN and distribution of candidacies also changed repeatedly. Initially, the party president personally appointed all secretarías in the CEN. In 1993, the party instituted a formal proportional representation system. This planilla system was later abolished at the party s sixth Congress in Finally, disloyal opposition, low levels of trust, and perceptions of illegitimacy almost routinely marred the use of the PRD s collective decision-making institutions. In particular, throughout the second half of the 1990s, PRD elites who competed with each other via closed primaries for major candidacies and top internal leadership positions repeatedly alleged fraud and refused to recognize the results. The early PRD s internal dysfunction, 187 This was one stated reason why Jorge Alcocer (PCM/PSUM/PMS) defected from the PRD in late Party elites agreed, for example, to hold a closed primary to determine the candidate for the 1997 Federal District mayoral election. 189 The party s first two presidents were elected by the first and second Congresses. Subsequently, the PRD membership voted for party president in closed primaries, independently of party congresses. 190 The number of secretarías also changed numerous times over the course of the 1990s. 245

254 frequently publicized in mass media, damaged the party s image among the broader electorate. Internal and external critics argued that the PRD, by modeling democracy so poorly internally, disqualified itself as a governing alternative and force for democracy in Mexico. 191 A document approved at the sixth Congress in 2001, for example, partially attributed the crushing defeat of Cuauhtémoc Cárdenas and the PRD in 2000 to a public loss of confidence in the PRD following the disorderly 1999 internal presidential election (about which more below). 192 Despite all of the above, the early PRD averted the worst. High levels of heterogeneity, weak horizontal linkages, fierce and constant internal battles over ideas and cuotas de poder, and weakly routinized procedures for resolving conflicts did not give rise to serious fragmentation or fatal schisms. What factor/s enabled the early PRD to avoid fragmentation and schism? The indispensability of Cuauhtémoc Cárdenas As Chapter One argues, two general factors decrease the likelihood of debilitating schisms, especially in large, heterogeneous parties with weak horizontal linkages and major divisions: (1) a strong electoral incentive for elites and factions to remain in the party, and (2) a mechanism for balancing internal interests and adjudicating internal conflicts. For almost all established political parties, the party label, or brand, provides the first. Many established parties, over time, institutionalize a set of procedures for adjudicating internal affairs, which provides the second. For the most part, new parties lack successful brands and institutionalized conflict resolution procedures. The early PRD was no exception. During its first decade, the party was still años del PRD: Porfirio Muñoz Ledo, May 22, See Prud homme (2003): 118, 118 note

255 attempting to establish a solid and far-reaching partisan brand in the electorate. It was also attempting and, as shown above, failing to institutionalize procedures for the balancing of diverse interests and the adjudication of disputes. Throughout the late 1980s and 1990s, party leader Cuauhtémoc Cárdenas substituted for a strong brand and also for a set of institutionalized conflict resolution procedures. Although the PRD was beset by internal conflicts over political tactics, cuotas de poder, and program during the formative phase, the Cárdenas brand provided a strong electoral incentive against lower elite defection. Equally, through active brokerage and arbitration, rooted in unquestioned internal authority, Cárdenas balanced internal interests, resolved conflicts, and enabled the party to act in a minimally unified manner. In these ways, Cárdenas played a critical role in preventing fragmentation and schism within the early PRD. Cárdenas s brand and coattails Not just any leader could have substituted for a strong left party brand in Mexico during the late 1980s. Cuauhtémoc Cárdenas, however, possessed tremendous, unrivaled electoral clout. One Mexico City-based journalist and political activist deftly observed that in contrast to the early PRD, which lacked access to electronic media and could not build support quickly, Cárdenas, from the beginning, did not need the media, or even a campaign. He communicated just in virtue of existing. With a name like Cárdenas, the campaign was carried out every day in school. 193 The genesis of the FDN and PRD demonstrated the fundamental importance of Cárdenas s electoral clout for left unity. Despite its extraparliamentary origins, the PRD required an electoral sphere...to become consolidated as a political force (7, emphasis added). The 193 Interview with Lajous. 247

256 Mexican left initially united, first in the FDN (1988) and subsequently in the PRD (1989), as a means to win the presidency: in the FDN to win the 1988 contest, and in the PRD to win the 1994 contest. In 1988, diverse left and center-left forces circled around Cárdenas upon recognizing his unique capacity to draw left and independent voters. As noted earlier, these forces included the PMS, which supported Cárdenas after it became clear that the entire left electorate, including traditional Marxist constituencies (e.g., UNAM students), supported Cárdenas s 1988 bid.no PRI elite save Cárdenas, writes Bruhn (1998), could...have united united all the center-left opposition parties or [reached] the agreement with the Mexican Socialist Party that gave the PRD institutional life (102). In the 1988 election, Cárdenas showed that he could win. In fact, his performance suggested that in 1994, with a stronger organization and improved defend-the-vote operation, he would win. 194 No other left leader could credibly offer the possibility of national power. As evidenced by his 1988 campaign and performance, 195 Cárdenas s personal electoral brand eclipsed that of all previous left parties and leaders. These facts cemented PMS and broader left support and provided the central impetus for the creation of the PRD. During the PRD s formative years, the PRD continued to [depend] heavily on...cárdenas to draw together diverse political allies (Bruhn 1998: 190). In early 1991, Arnoldo Martínez Verdugo (PMS) stated that there is a man whose level surpasses the PRD, who is a national leader... That is Cuauhtémoc Cárdenas. 196 In October of 1993, Heberto Castillo stated that there is no other citizen in the country with greater potential to unite currents around him than Cuauhtémoc. 197 Lower elites thus defected at their own electoral peril. Rodríguez 194 Interview with Córtez. 195 The convergence of Mexican left forces in the FDN (1988) and PRD (1989) was produced and consolidated by Cárdenas s [1988] candidacy and performance (Borjas 2003: 290). 196 Proceso, no. 741, January 14, Proceso, no. 874, June 14,

257 (n.d.) argues that even if Cárdenas had attempted to institutionalize his dominance within the early PRD by establishing a formal cardenista faction, a schism would not have resulted because there were few external incentives for the party s intermediate leaders to split (257). Coming out of the 1994 election, which he lost in a landslide, Cárdenas did not have the same proven external clout as after Yet the impact of this election on perredistas perceptions of Cárdenas s external clout should not be overstated. To begin, although Cárdenas suffered a crushing defeat, many PRD leaders and most of the base attributed his loss, primarily, to PRI spending and media manipulation, not to his weaknesses as a candidate. Moreover, Cárdenas resurged in 1997, winning the first ever election for mayor of the Federal District in a landslide. 198 In the view of nearly all perredistas, no other left leader stood a comparable chance of winning the 2000 presidential election. More than that, most believed Cárdenas stood a good chance of winning. Given his performance in the 1997 DF election and the PRD s subsequent gubernatorial victories, the Party of the Democratic Revolution approached the 2000 presidential elections with optimism (Borjas 2003, vol. II: 238). In sum, during the PRD s formative years, Cárdenas delivered the votes. The PRD brand, by and large, did not. To quote Borjas (2003), the charismatic leader and not the PRD was the receiver of the votes... The incorporation of citizens, organizations, and social movements resulted from his mobilizing capacity and not from a party program... (507-8). Cárdenas both was the PRD and was more than the party, since his leadership was recognized by citizens and non-activist groups willing to back him. In synthesis: the PRD was Cárdenas, but, surpassed by him, without his presence it was nothing (508). It should also be emphasized that as the 1990s unfolded, the PRD brand, as distinct from the Cárdenas brand, strengthened considerably. Surveys indicate that by 2000, most Mexicans percent of the vote to 25.6 (PRI) and 15.6 (PAN). 249

258 could locate the PRD on the left-right spectrum. 199 Whereas in the 1991 Chamber election, with Cárdenas off the ballot, the PRD had received only eight percent of the vote, in the 2003 Chamber election, the PRD s first election with Cárdenas off the ballot since 1991, 200 the party received eighteen percent of the vote. Thus, as Cárdenas s brand weakened over the course of the 1990s, the PRD brand strengthened, creating an offsetting electoral incentive against elite defection. Cárdenas s unquestioned internal authority In addition to substituting for a strong party brand, Cárdenas substituted for internal decision-making procedures by acting as the PRD s key decider and, in the event of conflict, its informal high court, or arbiter of last resort. Cárdenas s ability to assume these functions flowed from his unquestioned internal authority, rooted in a combination of factors: unmatched external appeal, which gave Cárdenas leverage over potential elite competitors, but also (1) high moral stature within the party, especially the base, and (2) a singular position, established very early, at the hub of the PRD s sprawling, heterogeneous, and fragmented vertical and horizontal networks. Cárdenas s internal moral stature Cárdenas s mística and moral stature within the PRD stemmed from several factors: his lineage, but also his political record and experiences prior to the PRD s creation and his intransigent opposition to the PRI throughout the PRD s formative phase. Although Cárdenas s lineage elevated him in the eyes of the broad electorate, as emphasized earlier, it especially 199 Citation needed. 200 In 1994 and 2000, Cárdenas ran for president, and in 1997, he ran for the Federal District mayoralty. 250

259 raised his stature within the left. Cárdenas was the last name most respected by the political left (Borjas 2003: 293). In the formulation of one Mexico-based political scientist, General Cárdenas symbolized the aspirations of the Mexican left, and the Mexican left associated Cárdenas the son with his father. 201 Cárdenas s lineage endowed him with an almost mystical quality for much of the PRD base. Among perredistas, Cárdenas symbolized the party s highest aspirations. The party and its moral leader were inextricably linked. PRD activists and supporters attributed to Cárdenas, more than to any other party leader, the authentic commitment and will necessary to bring about democratic change in Mexico. Cárdenas s actions and experiences during the 1980s helped solidify his stature. As governor of Michoacán (1980-6), he opposed and defied the neoliberal reforms of the de la Madrid government. More importantly, in 1988, according to nearly all Mexican leftists and left supporters at the time, Cárdenas unofficially defeated Salinas, sent the long-hegemonic PRI into an unprecendented internal crisis, and came close to toppling the regime. Subsequently, he refused to be purchased, rejecting Salinas de Gortari s offer of the regency of Mexico City and insisting on democratic intransigence and sustained confrontation. Because of these actions, and because of his name, Cuauhtémoc Cárdenas, at the PRD s founding, possessed unrivaled moral stature on the left. In the words of MUP leader Javier Hidalgo, Cárdenas had the moral quality to be everyone s leader. 202 Cárdenas s mística and moral stature carried the most weight at the base level. Cárdenas commanded the unconditional loyalty of most of the party rank-and-file, drawn primarily from the social left and the defecting PRI. Many of these base-level perredistas did not affiliate with a national corriente, did not hold office or paid party positions, and thus needed collective incentives in order to remain connected to the PRD organization (Rodríguez n.d.: 201 Interview with Somuano. 202 Interview with Hidalgo. 251

260 255). Cárdenas possessed an almost monopolistic capacity to produce collective goods for these groups and individuals (255). During the PRD s early development, Cárdenas shored up his support among the base by consistently prioritizing base-level demands, taking intransigent positions, and supporting the conception of the PRD as a movement, even at the expense of broader electoral appeal. On finding receptivity to their demands and avenues to promote them, the social left within the PRD became an interest group willing to obey, in any moment, the directives of the charismatic leader (Borjas 2003: 303) In short, just as the left-leaning electorate tended, during the late 1980s and early 1990s, to vote for Cárdenas rather than for the PRD, the early PRD rank-and-file was fundamentally loyal to Cárdenas rather than to the PRD. In Borjas s (2003) formulation, during the PRD s early years, the loyalty of members was directed fundamentally toward Cárdenas and only secondarily to the party itself (450-1). Many base-level perredistas identified the party with Cárdenas and considered the two inseparable. 203 Cárdenas s diehard support from the base, to some degree, reflected a broader trend. In general, Cárdenas commanded the most steadfast support among the more instransigent, confrontational sectors of the party. Although disproportionately represented at the base level, these sectors of the party also existed at the elite level. From the beginning, Cárdenas established alliances with a range of social left leaders with origins on the extraparliamentary left. Journalist and intellectual Arnoldo Córdova thus wrote in late 1990 that paradoxically, Cárdenas die-hards (incondicionales) come mainly from ultra-left groups, not from the CD. 204 Internal alliances between Cárdenas and the radical PRD leadership would last 203 From the perspective of Panebianco and given the importance conferred to Cardenas by the perredista bases, Cardenas arises as the creator and interpreter of a set of political symbols (the original ideological goals of the party) that come to be inseparable from his person (Borjas 2003: 450-1). 204 Unomásuno, Novemeber 16,

261 throughout the formative phase, in part because of Cárdenas s uncompromising stands as party leader. Thus, Cárdenas commanded the unconditional loyalty of the PRD base and certain radical sectors of the national leadership. In addition, given that he combined unmatched external electoral clout with unmatched internal, base-level support, Cárdenas enjoyed immense leverage over lower elite competitors and opponents. This leverage took three forms. First, lower PRD elites knew that they received extra votes due to Cárdenas s external appeal and coattails. Second, because Cárdenas commanded unrivaled loyalty among the PRD base, any lower elite who aspired to the party presidency, or needed broad internal support for some other reason, required Cárdenas s backing regardless of whether s/he agreed with Cárdenas s policies, tactics, and leadership style. In Borjas s (2003) formulation, PRD members loyalty to lower elites and members had an indirect character, given that [their loyalty] was fundamentally directed to [Cárdenas] and only in the second place to the party itself... The charismatic leader established himself as the source of other leaders legitimacy... (Borjas 2003: 450-1). Third, there was a broad recognition within the national PRD organization, among leaders of all stripes, that given the millions of external votes that Cárdenas delivered, and the countless individuals and organizations who joined and actively supported the PRD in primary allegiance to Cárdenas, PRD unity depended on Cárdenas. Cárdenas used both his loyalties at the base and elite levels and his leverage at the elite level to establish himself as the hub of the complex PRD complex organization. More specifically, Cárdenas established strong crossfactional ties both vertically (i.e., national to subnational) and horizontally (i.e., across national leaders and corrientes). Cárdenas s vertical crossfactional ties Cárdenas developed much stronger ties to the PRD s subnational leaders/organs than any 253

262 other national PRD elite. Not any leader could have achieved this, but Cárdenas, due to his intense, often die-hard support at the base level, established wide-ranging vertical ties with relative ease. Leaders and activists from the social left and ex-pri networks controlled most local and regional PRD offices. By and large, these individuals were cardenistas first and perredistas only by extension. Cuauhtémoc Cárdenas was the PRD, and the PRD was Cuauhtémoc Cárdenas. Consequently, organization and social movement leaders established personal, direct linkages with [Cárdenas], interfacing with the national party organization through him alone (Borjas 2003: 337). No other national leader claimed a remotely comparable network of subnational loyalties. The tight links between Cárdenas and the PRD s social left leaders meant that control...was shared and exercised almost completely by the group...headed by the charismatic leader (303). Subnational leaders across the PRD s territorial bastions displayed a willingness to cede [control] so that it could be assumed...by the charismatic leader...through bonds of loyalty established with him (450). The weakness of horizontal linkages across subnational organs further strengthened Cárdenas s position within the PRD organization. Most subnational organs engaged only in vertical communication, through Cárdenas, and did not form strong independent ties to each other. Rodríguez (n.d.) observes that Cárdenas s incontestable leadership became a node of organization between the different groups that...did not manage to organize horizontally (254, 263). Borjas (2003) similarly writes that the party s distinct subunits...more than interacting among themselves, established direct linkages with [Cárdenas] (448). Many internal critics and external analysts have claimed that Cárdenas intentionally monopolized vertical communications and deliberately undermined horizontal linkages across subnational organs and feeder groups in order to prevent the creation of independent alliances and counterweights within the party organization, and hence to maximize his internal control. Rodríguez (n.d.) states that Cárdenas reserved for himself negotiation with the local and 254

263 regional chiefs, establishing a network of political alliances under his control (254; emphasis added). Borjas (2003) argues that Cárdenas avoided the formation of horizontal linkages across leaders and across groups and guaranteed unconditional loyalties that permitted him to reinforce his control of the PRD (302-3). 205 Cárdenas s strong ties across the PRD s horizontally disconnected subnational organs gave him immense control in the critical zones of candidate selection, internal advancement and promotion, and programmatic and tactical development. Cárdenas used his vertical crossfactional ties to influence who entered the party, who ascended within the party appartus, who received important candidacies, what information lower organs received, and more. In exchange for loyalty and unconditional support from the movement and organization leaders who headed the PRD s subnational organs, Cárdenas listened to and communicated their grievances (Borjas 2003: 298) and awarded them candidacies and party posts (Borjas 2003: 337). 206 Through his strong ties to subnational leaders, Cárdenas also weighed in on local recruitment practices, helped determine which activists received promotions, and controlled the internal flow of information. 207 It should be noted that Cárdenas did not enjoy a complete monopoly on vertical ties. Rodríguez (n.d.) observes that the PRD s vertical structure of power, whereby subnational organs established individual, direct linkages with the national organization in pursuit of patronage, resources, and ideological influence, did empower other national leaders in addition to Cárdenas. Yet other national leaders vertical linkages paled in comparison to Cárdenas s, 205 See also Prud homme (1997: ); Martínez (2005: ), Rodríguez (n.d.: ). 206 The leaders of organizations and social movements established direct and personal links with the former presidential candidate and shared with him the role of executing the clientelistic exchanges within the organizations they represented. In this way, a complicated and solid network of relations arose within the PRD: social leaders gave loyalty and unconditional support to Crdenas, and in exchange they received candidacies and party positions (Borjas 2003: 337). 207 Cárdenas s control of information flows helped him impose the informal party line of democratic intransigence and sustained confrontation with the PRI (about which more below). See Borjas (2003):

264 who had relationships with a much larger number of local leaders than any other national elite, and who often commanded local leaders unconditional loyalty. Cárdenas s horizontal crossfactional ties Cárdenas supplemented vertical crossfactional ties with horizontal crossfactional ties at the national level. Just as he established hub-and-spokes relationships with the PRD s subnational leaders and organs, he established a network of direct, personal, and informal ties across the distinct national corrientes and their leaders. When Cárdenas skyrocketed to national prominence in the late 1980s, the establishment of direct contacts with [Cárdenas] became almost a rule for national left leaders (Prud homme 1997: 12, note 30). As early as 1988, many FDN and PMS elites had established direct ties to Cárdenas that supplanted their previous party disciplines and loyalties. 208 This dynamic persisted. Throughout the 1990s, Cuauhtémoc Cárdenas occupied the center, toward which the distinct corrientes with their respective leaders converged (Prud homme 2003: 118). Cárdenas position at the hub of the national organization stemmed, in part, from bonds of loyalty. Cárdenas commanded the respect and near unconditional loyalty of the intransigent sectors of the national leadership, drawn from social left organizations such as the ACNR, ORPC, and MAS. Yet other sectors of the national leadership did not possess similar loyalty to Cárdenas. In particular, national leaders from the traditional Marxist left and reformist sectors of the CD opposed Cárdenas s policies and often viewed him as a competitor. In order to draw these elites to the table, Cárdenas depended more on leverage than on loyalty. 208 Prud homme paraphrasing Jorge Alcocer. Prud homme paraphrases Alcocer that the establishment of direct contacts with [Cárdenas] became almost a rule. At the moment the PMS gave up its registry, the only nucleus that continued to function as a party organization was composed of ex-leaders of the PSUM and the PMT (12, note 30). Regarding FDN leaders, Borjas (2003) writes that they were aware, to varying degrees, that their position was the result of Cárdenas s vote, and as a result, each directly established bonds of loyalty and political commitment...to the charismatic leader, more than to the social or political organization that he or she came from (311). 256

265 Lower elites recognized Cárdenas s indispensability for the party s electoral success, for their own internal legitimation (should they need it), and for party unity. Consequently, they submitted to his leadership. One journalist and political activist from Mexico City independently invoked the hub-andspokes metaphor to characterize Cárdenas s dealings with lower elites in the national organization, or with individuals and groups who wished to join the PRD. Then, echoing the earlier critique that Cárdenas undermined horizontal linkages between subnational organs, the journalist argued that Cárdenas also undermined horizontal linkages within the national organization, intentionally meeting with only one or two people at a time in order to keep himself above others. 209 It should also be noted, however, that in order to stand above national leaders and corrientes during the PRD s formative period, Cárdenas made a consistent effort to avoid direct involvement in disputes and, when directly involved, not to dismiss any perspectives. Cárdenas himself wrote in August of 1995, I have tried to avoid being involved in disputes between groups and factions within the party. 210 At national congresses and other public debates, Cárdenas did not reject opposing points of view or lord his power over those who disagreed with him. Instead, in the public domain, Cuauhtémoc Cárdenas insisted on presenting himself as just another party member (Borjas 2003: 509) and attempted, where possible, to synthesize moderate and radical positions. * * * In summary, Cárdenas commanded the loyalty of the PRD base and radical elite; he enjoyed leverage over elite competitors due to his unrivaled combination of external clout and internal 209 Interview with Lajous. 210 Proceso, no. 982, August 28,

266 base-level support; and he occupied a singular position at the hub of the PRD s vertical and horizontal networks. National elites of an extraparliamentary bent supported him unconditionally. Reformist national elites from the PMS and CD recognized his indispensability for party unity, valued his coattails, and often, in pursuit of their internal objectives (e.g., the party presidency), could not afford to oppose him. Due to vertical and horizontal centralization in the person of Cárdenas, getting the party to adopt a policy depended less on debate or construction of a compromise than on convincing Cárdenas (Bruhn 1998: 190). Due to all these factors, Cárdenas stood above the different leaders that converged and coexisted in the PRD, 211 and every perredista knew it. In short, Cárdenas enjoyed unquestioned internal authority. Key illustrations of Cárdenas s unquestioned authority ( ) What did Cárdenas do with his unquestioned internal authority? In broad terms, as already stated, Cárdenas acted as the PRD s substitute for institutionalized procedures of decisionmaking and conflict resolution. Recognizing and capitalizing on his unquestioned internal authority and indispensability, Cárdenas, as he saw fit, imposed decisions unilaterally and without consulting party leaders or statutes. When interests required balancing, or conflicts arose, Cárdenas served as an active broker and arbiter, mediating and negotiating agreements and adjudicating internal conflicts. In order to manage his decision-making agenda and responsibilities as leader, Cárdenas, in a practice known as comunización, appointed a network of aides, to whom he delegated and entrusted different tasks. Although these individuals did not hold formal positions within the PRD apparatus, they spoke for Cárdenas and thus, along with Cárdenas, acted as the hub of the PRD s vertical network and national horizontal network. The cardenista nucleus 211 Borjas (2003):

267 never became a formal corriente, but it did constitute, for Prud homme (1997) and Rodríguez (n.d.), the early PRD s dominant coalition. 212 The central factor of this dominant coalition, Rodríguez (n.d.) writes, was Cárdenas s incontestable leadership (liderazgo incontestable) (256). What were the critical, contentious decisions that Cárdenas rendered or imposed during the PRD s formative years? Cárdenas exercised his power and vanquished internal competitors in three critical areas: candidate selection, internal promotion, and programmatic/tactical development. Within each of these areas, a number of key processes and episodes illustrate Cárdenas s enduring and near absolute control of the early PRD. The consolidation of the line of intransigence ( ) From 1989 to 1993, national conflict within the PRD centered, most centrally, on whether the PRD should embrace intransigence, mobilization, and the party-movement model (the intransigente, or rupturista view), or instead seek change through reform in the institutional sphere (the reformist view). To the chagrin of reformists, Cárdenas s informal decisionmaking and arbitration, during the late 1980s and early 1990s, wholly substituted for internal procedures and led decisively to the consolidation of the línea de intransigencia. At the PRD s founding Constituent Assembly of May 1989, national leaders agreed, almost universally, on the right to organize in corrientes. The PRD s initial corrientes corresponded to the party s feeder categories. The most disciplined corriente, the PMS, 213 supported reformist positions, as did the CD corriente s more moderate wing (headed by Porfirio Muñoz 212 Prud homme (1997) treats Cárdenas and his nucleus of loyalists as a hegemonic dominant coalition, while Rodríguez (n.d.: 256) treats them as a Bonapartist dominant coalition. Borjas (2003) treats Cárdenas as the core of a broader dominant coalition that also included Porfirio Muñoz Ledo and Heberto Castillo. 213 Martínez 2005: 61). 259

268 Ledo). The more radical sector of the CD, headed by Cárdenas, supported rupturista positions, as did the social left corrientes (e.g., ORPC, MAS, ACNR, OIR-LM). Reformist groups from the CD and PMS insisted on giving corrientes primacy, in part, to establish counterweights to Cárdenas s internal powers (facultades) (Martínez 2005: 101). 214 Yet party elites allowed Cuauhtémoc Cárdenas, the unofficial leader, to appoint all thirtytwo members of an interim CEN. 215 Tellingly, they also agreed that the CEN would not include a Secretary General position, which would strengthen the structure but counteract [Cárdenas s] personal leadership (Martínez 2005: 101). Explaining the decision not to include the Secretary General post, Andrés Manuel López Obrador stated in La Jornada: Competing power centers cannot be generated in the party; the indisputable leader is Cuauhtémoc Cárdenas. 216 Cárdenas s CEN appointments reproduced the balance of power between corrientes (Martínez 2005: 61), but also reflected Cárdenas s own personal ties. Composed of leaders who symbolized the party s struggle, and characterized by the coexistence of many partisan and organizational cultures (100), the first CEN included fifteen CD members, split between rupturistas and reformists, ten social left leaders, and six traditional Marxist leaders, all pemesistas (PMS). A large number of these individuals were Cárdenas s friends, previous collaborators, fellow ex-priístas, new allies, and fellow rupturistas (61, 63). The appointments produced instant recriminations from excluded groups (e.g., MAP) and quarrels within and between the CD and PMS corrientes, especially in Mexico City. 217 The 214 Porfirio Muñoz Ledo added in a 2009 interview that the corriente-based internal structure of the PRD reflected the party s PRI origins. We organized the party in corrientes, he stated, because we [the Corriente Democrática] were a PRI corriente. See 20 años del PRD: Porfirio Muñoz Ledo, May 22, Informal negotiations between corriente leaders determined the composition of the provisional Consejo Nacional. 216 La Jornada, May 7, Martínez (2005): 63, including note

269 infighting led Cárdenas to create a Coordinating Commission, composed mainly of allies, which became responsible for the party s fundamental decisions until the formation of the next CEN (Martínez 2005: 63). Not by chance, most individuals kept off the Cárdenasappointed Commission had supported reformist, not rupturista, positions in the previous months (Martínez 2005: 63, 100). At the Foro de Estatutos, held in July 1989, national leaders reviewed and finalized the basic statutes to be voted on at the first Congress in November This Congress would also elect the party s first full-term president and Consejo Nacional. The statutes agreed upon at the Foro granted extensive powers to the president, assumed in advance to be Cárdenas. Not only could the president appoint all members of the CEN, the CEN had to vote as a bloc, such that opposition corrientes in the CEN could not easily veto presidential initiatives. The Foro reaffirmed the removal of the Secretary General post, which consolidated [Cárdenas s] ascendance and role as a substitute for the lack of institutionalization (Martínez 2005: 101, emphasis added). 218 With these measures, Cárdenas, supported by the radical sector of the CD and groups on the social left, confirmed his authority in detriment to the factions interested in regulating his authority (mando) (Martínez 2005: 101). In addition to using appointment powers to strengthen the rupturista position in the national PRD organization, Cárdenas wielded his base-level clout to impose the line of intransigence on reformist elements at the subnational level. Rodríguez (n.d.) observes that Cárdenas s moral leadership...allowed him to control the terms of negotiation with outside political forces (255). He was the one who could establish the line separating friends from enemies, and the boundary between congruencia and betrayal, central axes of political identity, particularly visible in contexts of high political polarization such as the PRD s (255). Consequently, according to Borjas (2003), he could [impose] his vision of the PRD Consolida su ascendencia y papel supletorio a la falta de institucionalización. 261

270 as a party-movement and justify the line of democratic intransigence and the strategy of confrontation with the government (Borjas 2003: 451). 219 By way of illustration, after the mid-1989 municipal elections in Michoacán, PRD senator from Michoacán, Roberto Robles Garnica, criticized party activists for conducting prolonged occupations of various city halls in protest of perceived fraud. A group of local PRD legislators accepted the municipal results and participated in the formation of plural local governments. In response, Cárdenas issued a statement in support of the protesting activists. Despite Robles s public critique and the local PRD legislators acceptance of the electoral results, the protesters, on learning that they had the charismatic leader s support, maintained the occupation of the city halls (Borjas 2003, p. 352), and Robles and the local PRD legislators submitted to the line that Cárdenas imposed (Borjas 2003, p. 357). To observers, such episodes suggested an anarchy within the PRD that only Cuauhtémoc Cárdenas could counteract due to the arbitrating role that local leaders gave him (Borjas 2003: 371). Throughout this period, the top two lower elites, Porfirio Muñoz Ledo and Heberto Castillo, openly opposed two central developments: first, the consolidation of the line of intransigence; second, the reliance on Cárdenas s informal decisions, base-level appeals, brokerage, and arbitration, not on routinized procedures, for distributing positions and choosing party tactics and program (Borjas 2003: 303-6). Nevertheless, they possessed virtually no internal leverage and decided to back the charismatic leader (305-6). Neither Muñnoz Ledo nor Castillo and the traditional Marxist left played a relevant role in the decision-making process (306). Their proposals and critiques were debated, but they ended up being discarded democratically by the cardenista majority, which declared in favor of intransigence 219 Borjas (2003) also observes that control of communication determined that the flow of information occurred in a vertical manner, reinforcing the necessity of maintaining the party line of intransigencia democrática and the strategy of confrontation with the government (299). 262

271 and confrontation (Borjas 2003: 303-4). Equally, although Muñoz Ledo and Castillo enjoyed a margin of autonomy in certain spheres (e.g., Muñoz Ledo in the legislature), and also intended in the longer term to change the party s institutional and tactical profile, they recognized Cárdenas s initial and short-term importance as a unifier of diverse groups, a representative of the collective party identity, and a decisive source of unity necessary for party consolidation (305-6). At the PRD s first Congress, in November 1990, party radicals ascended. Delegates to the Congress, selected through different procedures by municipal and state organs, voted for the party presidency and the Consejo Nacional. In an election outcome that Martínez (2005: 64, 102) calls natural and a rubber stamp (tramite), Cárdenas ran for the party presidency unopposed and prevailed. The rupturista social left fared well in the concurrent election for the CN, with a plurality of twenty-three seats to the CD s twenty-one and the PMS s eighteen. Similar to Cárdenas s previous interim CEN, the new CEN s composition reproduced the balance of power expressed in the Congress with the additional ingredient of loyalty to the leader (Prud homme 1997: 15). At the Congress, Cárdenas unambiguously endorsed intransigencia, rejecting any negotiation that did not entail the dismantling of the PRI regime (64). Prominent traditional Marxists who had criticized the línea de intransigencia, notably Jorge Alcocer 220 and José Woldenberg, did not receive CEN appointments, prompting their defection and strong objections from other prominent traditional Marxists such as Heberto Castillo and Arnoldo Martínez Verdugo (Borjas 2003: 375). 221 The 1991 midterm elections, while a major setback for the PRD as a whole, strengthened the internal position of intransigentes. First, in internal, relative terms, the social left fared very well in the 1991 elections, electing eighteen federal deputies, to the traditional left s thirteen 220 Jorge Alcocer was a respected leader of the PCM/PSUM/PMS. 221 Castillo voted against the CEN, while Martínez Verdugo requested exclusion from the CEN (Borjas 2003: 375). In subsequent years, Martínez Verdugo would shift away from reformism in favor of intransigencia. 263

272 and the CD s four. In Prud homme s (1997) formulation, these results displaced the CD as the support base for the party leadership (Prud homme 1997: 16 note 41). Second, the election results did not undermine the authority of Cárdenas, who had not participated. In fact, for some, the results bolstered Cárdenas s authority, evidencing the PRD s inability to win votes without Cárdenas atop the ticket. 222 Cárdenas remained the source of unity of the groups that converged in the party and, moreover, through rhetoric and actions consistent with party objectives, he distributed the collective incentives necessary to maintain party activism and participation in spite of defeats (455). Even though pemesistas and CD moderates had fared poorly, they read the PRD s overall results as vindicating: at the ballot box, intransigencia evidently had not paid off. Heberto Castillo and other reformists criticized radical groups for their parallel clandestine activities and, as reported in Financiero, called on the PRD to force them out of the shadows. 223 Party radicals, for their part, attributed the PRD s poor results to fraud and other disadvantages and argued for continued intransigence. Cárdenas denounced the PRI s electoral abuses and, in order to sustain morale at the base level, intensified his rhetoric of intransigence. Radicalizing the battle of ideas in favor of confrontation with the PRI, writes Borjas (2003), constituted the only alternative for the PRD to guarantee its survival as a political force (455). In sum, during the PRD s initial few years, internal struggles centered, most immediately, on the intransigente/reformist debate, and Cárdenas imposed the line of intransigence. As the one responsible for building bridges, but also interested in dodging rules that would limit his leadership, Cárdenas personally staffed the early CENs and Coordinating Commission 222 It is important to highlight...that the authority of Cuauhtémoc Cárdenas was not questioned: not having directly participated in the contest, the defeat was the PRD s as an institution and not Cárdenas s (Borjas 2003: 416). 223 Statement by Ricardo Valero, Ricardo Pascoe, and Heberto Castillo in El Financiero, May 3,

273 and sought to reproduce the rough balance of power between corrientes (Martinez 2005: 99). Yet simultaneously, so far as possible, he appointed allies and tilted the balance toward fellow rupturistas. He imposed the line of intransigence, where necessary, on reformist party elites at the subnational level (e.g., Michoacán). His firm intransigencia and support from the PRD base contributed decisively to the ascendance of radical social left forces at the 1990 Congress and in the 1991 legislative elections. In all these ways, Cárdenas gradually steered the ship in favor of the social left and tactics of mobilization and confrontation (99). Through these actions, Cárdenas demonstrated that his informal role as decider and arbiter prevailed over any routinized procedures for collective decision-making and conflict resolution. Between 1989 and 1993, Martínez (2005) writes, What is the party and how is it going to work? were questions to which perredistas did not have a common answer, beyond their recognition of Cuauhtémoc Cárdenas s undisputed leadership (liderazo indiscutido) (104-5). Borjas (2003) summarizes that a few years into the PRD s existence, relations between PRD elites still emerged around specific disputes for positions that, arbitrated by Cardenas...reinforced the predominance of the charismatic leader, who continued imposing himself above institutional procedures (Borjas 2003: 516). In the early PRD, Cárdenas s word was effectively law. It should be noted that many early PRD elites, predominantly reformists, objected to the extreme and informal centralization of power in Cárdenas s person. Cárdenas s power, they charged, diminished the independence and vitality of local leaders and organizations, prevented the creation of horizontal linkages, and above all, prevented internal routinization. Jorge Alcocer, for example, emphasized in his resignation letter that crucial decisions took place through informal exchange and negotiation, always revolving around Cárdenas, instead of formal and transparent decision-making processes that would empower the party base. This disqualified the PRD, he suggested, from claiming the mantle of democracy: In order 265

274 to participate in discussion, one must be aligned with Cuauhtémoc Cárdenas. If authoritarianism is the enemy, the PRD must be consistent in its internal life. The truth is that party-related manners are not settled through the party s formal institutions but through private consultations [with Crdenas], an arrangement that concentrates functions in the party president (Proceso no. 740, 1/7/1991). Heberto Castillo quickly became marginal within the PRD due to his outspoken, frequently public critiques of Cárdenas s caudillismo (e.g., in Proceso). Castillo s public airing of grievances damaged his reputation among the large cardenista base of the PRD (Borjas 2003, vol. II: 448), and Castillo did not possess a strong independent base of his own. Even within the PMS, he mainly commanded the support of PMT cuadros. Although Muñoz Ledo agreed with Castillo that Cárdenas s leadership style undermined routinization, he made a realistic, calculated decision to continue [accepting] the predominance of Cárdenas s leadership (449). Only in subsequent years, when political circumstances changed, and he had accumulated significant political capital as a PRD legislator, would Muñoz Ledo attempt to challenge Cárdenas for leadership of the party. In late 1992, a PRD corriente, Corriente por la Reforma Democrática del PRD, formed in opposition to the authoritarian, personalistic structure of the PRD. Echoing a commonly voiced critique, 224 the corriente s founders argued that Cárdenas, instead of encouraging subnational organs to develop their own identities and voices, created vertical power structures based on personal loyalty. In a document published in Proceso, they wrote that the informal network of loyalties [to Cárdenas] constituted the real leadership of the PRD and was erod[ing] the organizational bases of the PRD and replacing a policy of activist 224 According to one party founder, MUP member, and early federal deputy, Cárdenas put loyalists in charge of state organs instead of encouraging them to develop their own identities and voices (interview with Flores). 266

275 formation and education with one of cooptation to loyalty. 225 Selecting the 1994 presidential candidate According to some perredistas, prior to the 1988 presidential election, Cárdenas and Porfirio Muñoz Ledo agreed that if Cárdenas lost in 1988, Muñoz Ledo would receive the candidacy in Cárdenas has denied this claim. 226 Regardless, the PRD did not have a formal procedure in place for deciding the candidacy. Effectively, the Mexican public and PRD base decided the matter by demanding that Cárdenas run. Because PRD members and supporters revered Cárdenas, and also recognized his unique capacity to attract support outside the party, they overwhelmingly pressed for his candidacy in The PRD base did not demonstrate its support for Cárdenas through formal procedures. As would occur six years later, support for Cárdenas s candidacy bubbled up outside formal party channels, from perredistas and non-perredistas alike. On February 5, 1993, at a public event celebrating the anniversary of the Mexican Constitution s signing, nearly one-thousand Mexican citizens, mostly but not exclusively PRD activists, proposed that Cárdenas run. Cárdenas provisionally accepted and later leveraged this internal and public support to secure the candidacy through informal negotiation. In a late 1990s interview, Muñoz Ledo stated that overwhelming internal demand for Cárdenas s candidacy induced him, in this informal negotiation, not to compete for the nomination: It s clear that Cuauhtémoc s leadership of the party was very broad, and that our people were...programmed for revenge... Overwhelmingly, the majority of the party were for [Cárdenas]... So I told him that I wouldn t compete Proceso no. 833, 10/19/ See Borjas (2003: 531 note 215) and Borjas (2003, vol. II: 420-1). 227 Proceso, no. 1155, December 20,

276 The 1993 election for president of the PRD On February 28 of 1993, Cárdenas resigned the presidency of the PRD to begin his campaign for the Mexican presidency (Martínez 2005: 67). Many national PRD leaders believed that Cárdenas s absence from internal party affairs might facilitate routinization. With Cárdenas on the campaign trail, the rest of the party leadership might have time and room to hammer out and consolidate formal decision-making and conflict resolution procedures, which would replace Cárdenas s informal arbitration (Borjas 2003: 508-9). In July 1993, several months after Cárdenas s resignation, the PRD held its second Congress. Like the first Congress, the second Congress would include the election, by congressional delegates, of the PRD s next president. The winner would replace interim president Roberto Robles Garnica, former priísta and member of the CD. The second Congress reflected and contributed to several important shifts in the early PRD. First, the election for party president triggered the first of several attempts, by the PRD during its formative phase, to compose the CEN via procedure, not through informal negotiations dependent on leaders ties and loyalty to Cárdenas. 228 The 1993 race for party president marked the first in which Cárdenas did not participate. Given Cárdenas s absence, the national party lacked its informal mechanism for distributing CEN secretarías postelection. Competing without Cárdenas s direct arbitration and aware of the risk of schism (fractura), the candidates for the PRD presidency agreed, in advance of the election, to use a formal proportional representation system in the allotment of CEN secretarías (Martínez 2005: 68). In particular, each candidate would run as part of a planilla, a formal coalition of national 228 The importance of the Third Congress resided in the fact that, for the first time, internal procedures were imposed over negotiations between different groups... [Leaders ] loyalty began to be directed toward the party itself and not just to Chauhtémoc Cárdenas (Borjas II 2003: 72). 268

277 leaders that could mix corrientes. A candidate s share of the congressional vote would determine the number of CEN secretarías and CN seats allotted to his or her planilla. The planillas would determine internally, via negotiation beforehand, which individuals would receive secretarías and CN seats. The agreement also stipulated that if the winning candidate lacked an absolute majority, the second-place planilla would receive the newly created number-two Secretary General position, the number-two post in the CEN. Second, the Congress both reflected and contributed to a shift in the axis of competition within the PRD. During Cárdenas s tenure as party president, corrientes and coalitions in the CEN and CN had corresponded very closely to the party s initial feeder organizations: the CD, the PMS, and different social left groups. By the second Congress, national leaders had reorganized the PRD s national coalitions around ideological proposals, some internally focused, others externally focused (Martínez 2005: 67). These new coalitions cut across and mixed the PRD s three initial feeder categories. Porfirio Muñoz Ledo, one of two PRD senators at the time, led the Arco Iris (Rainbow) planilla, which spanned the CD, PMS, and social left. The Arco Iris planilla favored a negotiated democratic transition. In addition, it called for the PRD to focus on a concrete governing program, externally, and on the routinization of decision-making and conflict resolution, internally. Three candidates ran in opposition to Muñoz Ledo. Heberto Castillo led the Cambio Democrático planilla, which included elements of the PMS and elements of the CD. Castillo based his campaign on a critique of the PRD s internal functioning: in particular, of the party s low transparency and its multiple layers of presidentialism, whereby each corriente followed the dictates of its leader, and the distribution of candidacies and internal positions across corrientes depended on opaque informal processes of negotiation and arbitration, all revolving around Cárdenas. His planilla proposed greater decision-making transparency and 269

278 routinization. Mario Saucedo of the ACNR led the Trisecta coalition, composed exclusively of social left and independent leaders. Trisecta opposed negotiation, embraced tactics of mobilization, and envisioned the PRD as an agent of social movements and organizations. Finally, Pablo Gómez of the PMS led a relatively small coalition that, like Castillo s, focused on internal issues. Gómez s planilla proposed a primary system for electing both party leaders and party candidates (Martínez 2005: 68-9). Although Cárdenas did not run in the election, he decided the outcome by providing critical, if unspoken, support to the bid of Muñoz Ledo. Whoever wanted to lead the PRD had to win Cárdenas favor (Borjas 2003: 522). Muñoz Ledo was the unquestioned leader of the PRD s legislative bloc and had demonstrated a unique capacity to reach deals outside of the party domain (Borjas 2003: 530). Given his reformist, institutionalist views and exceptional political abilities, most of the PRD s historical figures favored him (e.g., Ifigenia Martínez). Yet Muñoz Ledo had abstained from establishing ties of loyalty to the bases that might have allowed him to contest Cárdenas s leadership (Borjas 2003: 530). To win a plurality of congressional delegate votes, Muñoz Ledo needed Cárdenas s support. Although Cárdenas and Muñoz Ledo disagreed on fundamental questions of external tactics and internal procedure, Cárdenas quietly lent favor to Muñoz Ledo, 229 fellow CD member and the party s most able and decorated public servant. The party presidency constituted an implicit consolation prize for Muñoz Ledo, in exchange for Cardenas s having taken the PRD s nomination for the 1994 presidential race. 230 Cárdenas supported Muñoz Ledo for other reasons as well. Unlike Heberto Castillo, Muñoz Ledo had shied away from direct 229 See Borjas (2003): Muñoz Ledo himself later stated that he and Cárdenas struck a deal whereby Cárdenas took the presidential nomination, Muñoz Ledo the party presidency (20 años del PRD: Porfirio Muñoz Ledo, May 22, 2009). See also Borjas (2003):

279 criticisms of Cárdenas, even though Muñoz Ledo, like Castillo, objected to the personalization, centralization, and lack of routinization associated with Cárdenas s internal leadership (Borjas 2003: 530). In addition, Cárdenas trusted that Muñoz Ledo possessed the will and capacity to manage internal affairs effectively and lead the party to institutionalization, a prospect Cárdenas did not seem to oppose (Borjas 2003: 531). With Cárdenas s considerable weight behind him, Muñoz Ledo won with a large plurality of congressional votes (42 percent), while Saucedo placed second and Castillo a close third. Muñoz Ledo thus assumed the presidency and Saucedo the position of Secretary General. In order to marginalize Castillo, Muñoz Ledo named Pablo Gómez to the presidency of the Consejo Nacional (Pivron 1999: 263-4). All four candidates planillas received roughly proportional shares of CEN and CN seats. It should be noted that the second Congress, in addition to electing the new CEN and CN, approved and codified a new set of procedures for future internal elections. The new rules synthesized different corrientes proposals, mandating continued use of the planilla PR system and, crucially, a new primary system. In future elections for party president, candidates would have to compete for PRD members votes in closed primaries, not for delegate votes at party congresses. The 1994 crisis and Cárdenas s resilience The mid-1990s produced a set of challenges for Cárdenas s leadership. Even before the electoral failure of 1994, the PRD s internal battle of ideas had begun to shift in the reformists favor. The shift began with Muñoz Ledo s victory, facilitated by Cárdenas s backing, in the 1993 internal election. Although Cárdenas supported Muñoz Ledo s bid, Muñoz Ledo s victory, and the subsequent proportional allotment of CEN and CN seats, constituted an 271

280 advance for the moderate, pro-negotiation wing of the PRD. Whereas rupturistas had previously dominated the internal debate, after the 1993 internal, the disagreement between Muñoz Ledo and Cárdenas over the PRD s role in the transition to democracy became a less lopsided and more contentious battle within the PRD (Martínez 2005: 70). Between the second Congress and the 1994 presidential election, two sides would tighten the party : the rupturistas, composed of Cárdenas, CD sectors, and most of the social left, and the reformists, composed of CD sectors, the former PMS, and the former MAP (70-1). Tensions mounted between Cárdenas and Muñoz Ledo in particular. In a March 1994 interview with La Jornada, Cárdenas did not recognize the electoral reform agreements that Muñoz Ledo had reached with PRI legislators in the senate. Cárdenas s statement caused a rupture in the PRD congressional bloc, with almost half of the PRD s deputies voting against the agreement, and several abstaining (Martínez 2005: 70). In addition, Cárdenas openly questioned Muñoz Ledo s decision to fire and replace Samuel del Villar, PRD representative to the IFE. A CD member, intransigent, and Cárdenas appointee, Villar had caused controversy by raising questions about the Mexican electoral registry, controlled by the IFE (70). Reformists scored another important internal victory prior to the 1994 general election. Before the election, several members of the CEN left their posts in order to mount campaigns. Muñoz Ledo capitalized, using the occasion to reorganize the CEN and empower the reformist wing of the party. Muñoz Ledo distributed fourteen of twenty CEN positions to reformists, including the most important positions under his control. Nevertheless, before the 1994 election, the rupturistas still retained the upper hand. Although reformists formally controlled the internal apparatus after the July 1993 Congress, the number-two post of Secretary General still belonged to rupturista Mario Saucedo (ex- ACNR), and more importantly, the debate remained tilted toward the rupturistas given 272

281 the [party s] warlike relationship with Carlos Salinas and the strong leadership of Cárdenas (Martínez 2005: 70-1). Reformists thus had little choice but to accept Cárdenas s decisions, popular among the PRD s largely social left base, to meet with the Subcomandante Marcos of the Zapatista National Liberation Army, to oppose NAFTA, and to back out of negotiations on electoral reform. The 1994 election significantly affected the balance of the rupturista/reformist debate. Cárdenas underperformed abysmally, and in contrast to the 1991 congressional elections, moderates within the national leadership, in 1994, could read the electoral outcome as a public verdict against Cárdenas. The election, they argued, constituted an implicit rebuke of the línea de intransigencia that Cárdenas and his base-level supporters espoused. Rupturistas uncompromising stand on negotiation, in addition to other questionable choices such as meeting with Comandante Marcos and opposing NAFTA, had clearly not paid off electorally. With electoral results to back up their assertions, reformist critics and their arguments now carried more internal weight than before. After the 1994 election, instead of taking a formal position within the PRD, Cárdenas founded a think tank, the Fundación para la Democracia, and officially received the designation of moral leader of the PRD. The ascendance of the reformist wing culminated in events at the PRD s third Congress, in late August of The rupturistas, led by Cárdenas, proposed that the PRD demand Zedillo s resignation. The reformists, or dialoguistas, led by Muñoz Ledo, proposed that the PRD recognize Zedillo s legitimacy and participate in a pacted transition to democracy (Martínez 2005: 73). In May of that year, Muñoz Ledo had captured the prevailing view among dialoguistas, stating that it was necessary to decide if what we want is the defense of certain principles even at the risk of becoming marginal. 231 The Congress formally resolved in favor of the reformists. The rupturistas, after propos- 231 Proceso, no. 964, April 24,

282 ing a societal referendum on Zedillo s impeachment and losing another congressional vote, accepted the reformists victory. Despite the results of the Congress, Cárdenas s leadership and authority remained uncompromised. On the one hand, his persistence in the radical line continued to sustain the loyalties of the many perredista militantes who favored intransigence (Borjas 2003, vol. II: 64). On the other, Cárdenas took actions that strengthened his reputation as the PRD s internal peacemaker and unifier. First, he publicly recognized and praised Muñoz Ledo for his leadership and efforts (Martínez 2005: 74, note 56). Second, he openly addressed a prominent critique, apologizing to the party for having created an independent campaign committee during the 1994 presidential race. According to Pivron (1999), the recovery of Cárdenas s leadership reached its apex when he recognized his errors... By conceding, Cárdenas recovered his image as a great conciliator, a symbol of unity...who prioritized institutional over personal interests. In reality, he had regained (recobrado) his leadership (266). In short, coming out of the 1994 presidential election and the third PRD Congress in 1995, Cárdenas maintained broad internal authority. He retained his image as above faction and symbolic of the party as a whole, while his continued commitment to intransigence kept radicals psychologically invested in the PRD. 232 The period marked the apex of Muñoz Ledo s trajectory within the PRD. 233 Not only did he oversee the reformists victory at the 1995 Congress, he also played an instrumental 232 Bruhn (1998) summarizes that after 1994, Cárdenas remained the PRD s strongest unifying force and an important reference point for many perredistas, giving him latitude to affect party policy against the preferences of the national leadership by making public policy statements, ingenuously identifying himself as just a PRD member with liberty to express his opinion, but fully aware of his impact (Bruhn 1998: 193). 233 Borjas (2003, vol. II) summarizes Muñoz Ledo, in contrast to Cárdenas, as follows: Cárdenas s behavior was guided by the principles on which internal legitimization in the PRD was based... [If] he was the charismatic leader who moved the bases and disciplined the party leaders and legislative bloc, Muñoz Ledo was the statesman who, fortified by the pressure that that could represent, had managed to convert PRD principles into facts and concrete content (note 244, 121). 274

283 role in the passage of new internal procedural reforms 234 and, even more importantly, the landmark electoral reforms of 1996 in the Mexican national legislature. As noted earlier, these reforms represented a decisive step in Mexico s democratic transition, and Muñoz Ledo, among perredistas, could rightfully claim the most responsibility for them. Together, Muñoz Ledo s legislative activities and achievements consolidated the PRD s internal strategic shift from intransigencia to limited cooperation (Prud homme 2003: 103). Despite all this, Muñoz Ledo, as a leader, remained clearly secondary to Cárdenas. In mid- 1996, Muñoz Ledo launched a series of accusations against Cárdenas in order according to Borjas (2003: 86) to strengthen his position within the party. In the weeks before the mid-july election for party president, Muñoz Ledo stated publicly that Cárdenas, after the 1988 presidential election, had held two secret meetings with Carlos Salinas de Gortari and Manuel Camacho, the Federal District s newly appointed regente. In Proceso, Muñoz Ledo described these meetings as incongruous with Cárdenas subsequent stance of intransigencia. He also criticized the timing, questioning Cárdenas s decision to engage in dialogue when the public would have supported intransigence (i.e., in 1988), and later to insist on intransigence in a radical, strident, and unnecessary manner well into the 1990s, after the appropriate moment had passed. 235 Cárdenas denied that the meetings took place and called Muñoz Ledo s accusations lies and fantasies. 236 Muñoz Ledo s gambit backfired. Top national PRD leaders, as well as influential leftists not in the PRD, either backed Cárdenas (Heberto Castillo, Gilberto Rincón Gallardo, Jorge Alcocer) or remained neutral and downplayed the conflict (AMLO, Amalia García). 237 Jorge 234 The procedural reforms decentralized power in the national PRD, taking CEN appointment power away from the party president and putting it in the hands of the CN and, by extension, the corrientes and related planillas. See Martínez (2005: 74); Wuhs (2008: 56). 235 Proceso, no. 1026, June 30, Proceso, no. 1026, June 30, Borjas (2003, vol. II):

284 Alcocer accused Muñoz Ledo of sowing seeds of discord and suspicion. 238 The PRD base also forcefully supported Cárdenas. Thus, instead of becoming stronger within the PRD, Muñoz Ledo brought about the forceful confirmation and unconditional backing of Cárdenas s leadership by the PRD base and leadership (Borjas 2003, vol. II: 86). In the heat of the conflict, and just a week before the mid-july internal presidential election, Heberto Castillo wrote of Muñoz Ledo in Proceso: Porfirio never managed to escape from the shadow of the caudillo that Cardenas has been, first as the unity candidate of nearly all of [the country s] left forces, and later in the PRD, which is founded on the directives that [Cárdenas] imposes with his charismatic force among the perredista bases. 239 The 1996 election for PRD president In mid-july of 1996, over 300,000 perredistas, roughly a fourth of the party membership, cast their vote in a closed primary for Muñoz Ledo s successor. The 1996 internal election marked an important and enduring shift in the structure of competition within the PRD. While internal coalitions had corresponded to the party s initial feeder categories in 1990, and to ideological positions in 1993, in the 1996 campaign, internal coalitions mixed feeder groups and ideologies and reflected, above all, personal loyalties (Martínez 2005: 24, 214). This trend has persisted to the present day. PRD members elected Andrés Manuel López Obrador by an overwhelming margin. Amalia García, whom Muñoz Ledo backed, received under ten percent of the vote. Two thirds of the vote were cast in Michoacán, the Federal District, Mexico state, and AMLO s native Tabasco. In a well-calibrated campaign that targeted the PRD s organized bases, AMLO identified himself as center-left but also declared himself a revolutionary nationalist à la Cárdenas. 238 Proceso, no. 1026, June 30, Proceso no. 1027, July 7,

285 Most centrally, however, he placed emphasis on political tactics specifically, the importance of mass mobilization rather than specific policy proposals. AMLO characterized the party v. party-movement debate as a false choice: As some negotiate, others help by mobilizing society. 240 If the PRI committed fraud even at the gubernatorial level, he claimed, the PRD must use its mobilizing capacity to create a national crisis (Borjas 2003, vol. II: 82). He also insisted on negotiating with the PRI as equals, not as ants to elephants (Borjas 2003, vol. II: 82). AMLO s rise within the party did not result from internal moral stature, as in the case of Cárdenas, or from decades of accumulated political capital and connections, as in the case of Muñoz Ledo (Martínez 2005: 200-1). Instead, his rise resulted from short-term success at building a large coalition that cut across feeder categories and ideological perspectives. Crucially, AMLO s success at internal coalition-building stemmed from his proximity to Cuauhtémoc Cárdenas. 241 Cárdenas s backing...activate[d] bridges between previously opposed groups (Martínez 2005: 74). More specifically, in negotiations facilitated by Cárdenas, AMLO managed to strike deals and form electoral alliances with Mario Saucedo s radical corriente and Jesús Ortega s moderate Nueva Izquierda corriente. Prior to the coalition s formation, Saucedo had campaigned against AMLO s center-leftism (Borjas 2003, vol. II: 79), while Ortega, like fellow competitors Amalia García and Heberto Castillo, had opposed AMLO s emphasis on mobilization and argued that the PRD, in order to win at the national level, must abandon the logic of a political movement and adopt the logic of a political party. 242 Ortega and Saucedo decided to join forces with AMLO, recognizing the likelihood of an absolute majority for AMLO atop such a broad coalition. The leaders agreed that in the event of an absolute majority, the Secretary General position would go to Ortega, 240 AMLO s initial campaign speech, quoted in Borjas (2003, vol. II): Borjas (2003, vol. II) uses the term proximidad to describe the Cárdenas/AMLO relationship. 242 Proceso, no. 116, April 22,

286 while Saucedo s corriente would benefit in the proportional allotment of CN seats. Thus, in AMLO s new CEN, Ortega received the post of Secretary General. Other top members of AMLO s planilla possessed strong ties to Cárdenas, separately reflecting the influence of the PRD s undisputed leader on the new party president. 243 The 1997 DF election, the 1998 Congress, and Cárdenas s resurgence After the 1996 internal election, the internal debate between reformists and intransigentes remained intense and unresolved, and tensions between Muñoz Ledo and Cárdenas continued to mount (Borjas 2003, vol. II: 84). The electoral reforms of 1996 opened the Federal District to direct elections, raising the question who in the PRD would run. Both Cárdenas and Muñoz Ledo sought the party s nomination, and the party held a closed primary to determine the victor. [I]n terms of leadership and popular acceptance, the battle was lost in advance for Muñoz Ledo, given Cárdenas s clout with the PRD base (Borjas 2003, vol. II: 121). 244 After losing the internal, Muñoz Ledo accepted a candidacy for federal deputy. According to Borjas (2003, vol. II), Cárdenas s landslide victory in the first ever direct election for Federal District mayor elevated his internal status to unprecedented heights (233). To begin, Cárdenas benefited from the inevitable glow of victory. Further, his performance restored a sense of hope and confidence in perredistas that the party could win major elections. 245 Crucially, however, the July 1997 victory also elevated Cárdenas s status because 243 These included Adolfo Gilly, Rodolfo González Guevara, Carlos Lavore, and Ricardo Pascoe (Borjas 2003, vol. II: 83). 244 In a 2009 interview, Muñoz Ledo called the 1997 internal elections unfair and very doubtful, claiming that Cárdenas s supporters took over polling stations, and the party distributed campaign resources unequally. Muñoz Ledo claims to have informed AMLO, the party president, that he would not again tolerate that kind of internal manipulation (20 años del PRD: Porfirio Muñoz Ledo, May 22, 2009). 245 According to Borjas (2003, vol. II), Cárdenas s victory not only showed that democratic intransigence had borne fruit, but also that it was possible to continue reaping victories (233). 278

287 he had refused to make ideological compromises during the campaign. On the contrary, he had remained firmly committed to the línea de intransigencia. Cárdenas consolidated his elevated status at the PRD s fourth Congress in March Given the events of previous months, his leadership was never more complete than at the fourth Congress (Borjas 2003, vol. II: 233). Moreover, during the Congress, Cárdenas effectively played the role of bridge-builder. A central debate at the Congress concerned whether to define the PRD as a party of the left. As documented in Proceso, Cárdenas weighed in, stating that instead of using the terms right and left, he preferred to identify the party in terms of ideological content, particularly national sovereignty, liberty, rights of the masses, equality, distributive justice, universal access to education and health. 246 Borjas (2003, vol. II) writes that by separating himself from the PRD, Cárdenas once again placed himself above the party, given that he demonstrated a more inclusive attitude that allowed him to build bridges toward other political expressions and set himself up as the center and guarantor of any alliance (237). Cárdenas v. Muñoz Ledo: the selection of the PRD s 2000 presidential candidate Internal jockeying for the PRD s 2000 presidential candidacy began in late 1998, with Cárdenas in the middle of his mayoral term, and Muñoz Ledo nearing the end of his term as president of the Chamber of Deputies. Muńoz Ledo publicly revealed his intention to compete for the PRD s nomination in October of In late December, he made a case for himself in Proceso. While recognizing and lauding Cárdenas s firm ideological convictions, Muñoz Ledo suggested that the argument for Cárdenas rested mainly on lineage, and that the appropriateness of a third Cárdenas run would depend on the success of his hitherto 246 Proceso no. 1145, 10/11/

288 underwhelming tenure as mayor of the Federal District. 247 The following week, battle lines were drawn. Cárdenas, for the first time, stated that he was inclined to seek the PRD s nomination and cited the will of the PRD militancia as the critical determinant. 248 Everything depended on which side the militancia favored, and Cárdenas and Muñoz Ledo knew well which side [the militancia] would be on. Thus, as always, the moral leader of the PRD had the last word (Borjas 2003, vol. II: 290). In the early months of 1999, PRD internal leaders, legislators, and militantes rallied around Cárdenas, setting into motion of a series of events that eventuated in Cárdenas s nomination and Muñoz Ledo s exit from the party. In January of 1999, over one-hundred PRD deputies penned an open letter in support of Cárdenas s candidacy, which leaked after party president López Obrador, on Muñoz Ledo s entreaty, prohibited its circulation (Borjas 2003, vol. II: 290-1). Having previously called a truce in order to avert debilitating internal conflicts, Muńoz Ledo went on the offensive. He rejected the proposal to nominate the party s candidate through an internal primary (comicios internos), characterizing the proposal as a veiled maneuver for securing Cárdenas s nomination (292). In early April, after Cárdenas admitted to having met with Carlos Salinas de Gortari after the 1988 election, Muñoz Ledo excoriated the leader for inconsistency and duplicity and, in Proceso on the same day that Cárdenas formally announced his candidacy, called the revelation the collapse of a myth. 249 Muñoz Ledo s offensive maneuver only worsened his internal position. A growing number of PRD leaders, including many key leaders who had long sided with Muñoz Ledo on the party s central debates reformism v. rupturismo, institutionalization v. caudillismo closed ranks around Cárdenas and criticized Muñoz Ledo for doing detriment to the party. Foremost 247 Proceso, no. 1155, December 20, Proceso, no. 1156, December 27, Proceso, no. 1172, April 18,

289 among these was Amalia García, a reformist who had supported Muñoz Ledo s bid for the party presidency in 1993, and whose bid for the presidency Muñoz Ledo had supported in In May, the PRD-aligned wing of El Barzón, a popular business and farming movement, publicly declared their support for Cárdenas, while the PRD s State Committee in Michoacán went ahead and established a pro-cárdenas campaign committee (295). PAN overtures regarding a possible PAN/PRD alliance heightened Muñoz Ledo s vulnerability, as Cárdenas, in virtue of his proven electoral clout, would likely receive more support from panistas (294-5). At the end of May, Cárdenas imposed his candidacy in an unexpected manner. In what Muñoz Ledo would later call a disgraceful violation of party statute, Cárdenas, without formally consulting the party, accepted the nomination of the Partido del Trabajo, correctly calculating that this action would induce the PRD to support his bid immediately and, in this way, streamline the nomination process. Given that the PAN s Vicente Fox had been campaigning for the better part of the year, and that the PRI had also started the campaign season early, 251 speed mattered. Muñoz Ledo blasted Cárdenas and demanded that the CEN revoke Cárdenas s membership for violation of party statutes. Predictably, the CEN refused Muñoz Ledo s demand. Interim president Pablo Gómez said that to do so would be to confront Cárdenas directly, while to deny the soundness of Muñoz Ledo s argument would be to support Cárdenas. Both, he stated, were unacceptable, as the party leadership must remain neutral (Borjas 2003, vol. II: 297). Of course, by remaining neutral, the CEN definitively opened the way for Cárdenas s nomination a point not lost on Muñoz Ledo Borjas (2003, vol. II): As an aspirant to the party presidency in 1999, García and other moderates (e.g., Jesús Ortega) had an incentive to support Cárdenas s third bid (Martínez 2005: 80). 251 Quotation from Amalia García in Proceso, no. 1188, August 8, Proceso, no. 1184, July 11,

290 The selection of the PRD s presidential candidate in 2000 evidenced that, more than a decade after the PRD s creation, Cárdenas remained above any formal party statute. To summarize the above, in early 1999, Cárdenas almost secured the PRD s nomination informally due to the sustained outpouring of support he received upon announcing his tentative intention to run. Yet Muñoz Ledo doggedly pursued the nomination, and Cárdenas, if he had chosen to resolve the internal struggle in adherence to party statutes, would have had to engage in internal negotiation and perhaps submit to some type of internal election. This process carried an outside risk of losing and would have sucked valuable time and energy from the presidential campaign. Hence, instead of resolving the struggle institutionally, Cárdenas resolved it unilaterally and in violation of PRD statute, accepting the PT s nomination, forcing the PRD to nominate him immediately, and marginalizing Muñoz Ledo in a naked demonstration of power. 253 The message, Muñoz Ledo stated in Proceso, is that the PRD has an owner. 254 Almost a year later, Muñoz Ledo, no longer a perredista, would summarize: When there s someone above everyone, there is no law. We could never institutionalize the party. 255 The 1999 election for PRD president Mere weeks after the CEN s refusal to remove or even censure Cárdenas, the PRD held a closed primary to determine AMLO s successor. This primary caused the most severe internal crisis in PRD history. Four candidates ran, two moderates from the traditional Marxist left (Amalia García, Jesús Ortega), two radicals from the social left (Mario Saucedo, Rosa Albina Garavito). No ex-priístas ran. García and Ortega violated party statute in running, having 253 One more time, writes Borjas (2003, vol. II), the PRD s charismatic leader based his political leadership on external alliances in order to situate himself, in a convincing manner, atop Muñoz Ledo, but above all atop the PRD itself (295). 254 Proceso, no. 1178, May 30, Proceso, no. 1227, May 7,

291 contested the previous internal. In the official tally, García prevailed by a very narrow margin, and Ortega s supporters disputed the results, leading to the invalidation of the election and creation of an interim CEN. In the interim, all four candidates joined in a shaky coalition, with García atop the ticket. In the event of an absolute majority, Ortega s ally Jesús Zambrano would receive the Secretary General position, Albina Garavito would receive the position of consejera nacional, and all four candidates previous planillas would receive a share of the new coalition planilla s CEN quota. Eight additional candidates ran, including Ifigenia Martínez, backed by Porfirio Muñoz Ledo and his planilla, Nueva República. Cárdenas supported García, presumably in order to contain Muñoz Ledo. 256 García also claimed the support of AMLO and Ricardo Monreal, capable of delivering Tabasco and Zacatecas virtually wholesale (Martínez 2005: 83). García prevailed, and Ifigenia Martínez fared so poorly that Muñoz Ledo s Nueva República planilla won just one seat on the CEN. Utterly marginalized, Muñoz Ledo, in late August, accepted the presidential nomination of the PARM and, in January of 2000, formally exited the PRD (Borjas 2003, vol. II: ). 257 Martínez (2005) writes that, a full decade after the PRD s creation, the 1999 internal evidenced an alarming inability to channel conflicts through stable procedures (83). The 2000 presidential campaign evidenced the CEN s weakness, as Cárdenas s independent committee held primary decision-making power, and García s unsteady performance provoked internal criticism (Martínez 2005: 83). Following Cárdenas s landslide loss, the PRD called its sixth Congress in twelve years. Prior to the Congress, a conflict arose in the Consejo Nacional between one bloc, led by consejera nacional Rosa Albina Garavito, calling for 256 Martínez (2005): 83, note años del PRD: Porfirio Muñoz Ledo, May 22,

292 García s resignation, and another bloc, led by supporters of García and Ortega, supporting the completion of García s term. After García refused to step down, Albina, who had joined García s coalition planilla little more than a year earlier, resigned as consejera nacional. The Congress provoked a series of debates on tactics and program, with new wings forming around García/Ortega on the moderate, institutionalist side and Cárdenas along with ally Rosario Robles on the more radical, movimientista side. The Congress ended with collective agreement on the need for greater internal routinization but also, somewhat ironically, a new set of rules for distributing internal positions. The Congress abolished the planilla system, such that candidates for party president, moving forward, would run accompanied only by a candidate for Secretary General. 258 * * * The PRD thus remained weakly routinized throughout the formative phase. Until 1993, many basic formal procedures (i.e., for composing the CEN) still did not exist. Subsequently, despite various attempts to routinize decision-making and conflict resolution within the national organization, party leaders, including Cárdenas, regularly violated, rejected, and changed internal procedures. At the turn of the millennium, PRD s permanent need to revert to methods of informal arbitration in cases of grave internal conflicts evidenced the deficiencies and instability of the PRD s formal procedures (Prud homme 2003: 118, 104). Cárdenas s informal role as decision-maker and arbiter prevailed over any formal procedure. 259 Cárdenas concentrated an incredible amount of decision-making authority. Capitalizing on his moral stature, leverage, and singular position at the hub of the PRD s vertical 258 For details and qualifications, see Martínez (2005): Scholars converge in this assessment. The early PRD was led and arbitrated by the charismatic leader (Borjas 2003: ). On Prud homme s (2003) summary, Cárdenas exercised an important function of informal arbitration in the internal life of the organization (118). Martínez (2005) identifies three authentic birthmarks of the PRD. The first is Cárdenas s role as arbiter : The presence of Cárdenas as the leader of a movement broader than the PRD, who had great legitimacy and was placed at the 284

293 and horizontal networks, Cárdenas, during the PRD s formative phase, repeatedly swayed internal presidential elections, brokered the allocation of positions (e.g., secretarías) and candidacies, imposed himself as the party s candidate in major executive elections, and served as a powerful spokesperson, arbiter, and broker in the PRD s tactical and programmatic development. Although Cárdenas played a less active role in internal party life from the mid-1990s onward, internal procedures did not become consolidated, and Cárdenas remained able and willing to broker alliances (e.g., AMLO s 1996 coalition), tilt or sway internal presidential elections (1996, 1999), and impose himself as the PRD s top executive candidate (2000). Cárdenas often made these decisions without consulting other key actors. 260 Throughout the formative period, most moderate party elites and many outside observers criticized Cárdenas s leadership style, citing its negative consequences for routinization. 261 These critiques persisted throughout the 1990s. In late December of 1998, for example, perredista and former student leader, Raúl Álvarez Garín, publicly rebuked Cárdenas for his caudillista practices as party leader and jefe de gobierno of the Federal District: [Cárdenas] doesn t discuss problems. He receives them and considers them in private. He doesn t interact in debates. He doesn t risk making judgments or proposing solutions to certain problems in a daring manner. [His] hermetism, separated from debate, is damaging for collective political life. He establishes bilateral relationships, and this gives the appearance of favoritism or, even worse, solutions caused by particularistic influences. Naturally, due to his weight in the party and his modus operandi, these are practices that many consider unacceptable and are seen as caudillismo. 262 center of the party as an arbiter, established balance between forces that compete[d] for his support (97). The second birthmark is that the corrientes constituted the only avenue of participation. The third is weak routinization and a reliance on informal, short-term measures to resolve disputes. He later summarizes the early PRD national organization as divided between a reformist, traditional wing and a radical, extraparliamentary wing with a charismatic leader, acting as the cement, [who] seeks to balance disputes (101). 260 As Borjas (2003) summarizes, Cárdenas s charisma equipped him to impose on the party all key decisions without having to negotiate with the rest of the actors (Borjas 2003: 451). 261 Martinez (2005): Proceso no. 1155, 12/20/

294 Yet it could be argued that PRD unity, especially in the initial years, hinged on Cárdenas s assumption of decision-making and conflict resolution responsibilities. By substituting for an internal decision-making and conflict resolution procedure, Cárdenas enabled the PRD to act as one and speak in a single voice. Rodríguez (n.d.) suggests that given the PRD s unstable equilibrium, if Cárdenas had not acted as the party s informal decision-making center and high court, and instead attempted to routinize his charisma by creating a corriente cardenista and submitting to established procedures, the PRD might have fractured (256-7). Later, she writes: Diversity coexisted in a conflictual manner within the party, but during its initial years, [the PRD] generally managed to act in a unified manner due to Cárdenas, whose dominance...was sustained thanks to agreements with different groups and the recognition of his incontestable leadership (Rodríguez n.d.: 303). In a 2010 historical reflection, party founder and leader Jesús Ortega wrote: In our party, guevaristas, castristas, Christians, atheists, socialists, social democrats, liberals, and communists came together, and what kept them united? The strength and authority of Cuauhtémoc Cárdenas... (Ortega 2010: 18). 263 To this day, the PRD still has not become internally institutionalized. As late as 2005, perredista attempts to improve procedures remained undermined by the precariousness of their achievements (Martínez 2005: 88). In the last decade, party leaders have continued to resolve power struggles informally, and candidate selection procedures still change regularly. On multiple occasions, the PRD has required the intervention of outside bodies to adjudicate claims of internal electoral fraud. Yet since the late 1990s, weak routinization has posed less of a threat to the PRD s basic integrity, given the party s increased access to state patronage and financial resources. The 263 Ortega also cites the strength and authority of Andrés Manuel López Obrador. I exclude this part of the quotation because the chapter concerns the role of leadership during the PRD s formative years. AMLO s status as the PRD s leader only became established after Cárdenas s retirement from politics, which followed the 2000 presidential election. 286

295 party has continuously controlled the Federal District since 1997 and won over a dozen gubernatorial elections since The party has maintained a relatively strong presence in the national legislature, which, since the 1996 political reforms, has resulted in generous outlays of public funds. The PRD s hardcore vote (voto duro), combined with the spoils that now go along with it, provide much stronger incentives against elite defection than existed during the bulk of the PRD s formative phase. Perredistas continue to wage bitter and disorderly power struggles, but few dare to exit. 287

296 Chapter 6 Electoral Collapse: the Case of Argentina s FREPASO In their origins, Argentina s FREPASO and Mexico s PRD bear a significant resemblance. 1 Like Mexico s hegemonic PRI in the early 1980s, Argentina s dominant, traditionally populist Partido Justicialista (PJ) adopted major neoliberal reforms in the late 1980s. The Carlos Menem (PJ) administration took office amid economic crisis in 1989 and quickly lowered protectionist tariffs, carried out a battery of privatizations, and implemented strict, antiinflationary monetary policies. Menem s economic policies provoked opposition among leftleaning PJ deputies, eight of whom defected from the party in mid For the next-half decade, this Group of Eight (Grupo de los Ocho), in alliance with other left and center actors, created a succession of opposition fronts and parties, culminating in the formation of the Big Front (Frente Grande, or FG) in 1993 and FREPASO (Frente País Solidario) in late FREPASO quickly became Argentina s third electoral force, behind the PJ and UCR (Unión Cívica Radical). Like Mexico s PRD, FG/FREPASO thus originated as a left-wing, anti-neoliberal splinter group from a dominant, traditionally populist party, developed over the course of the 1990s, and became the first new party in decades to challenge the national two-party establishment. Four years after FREPASO s founding, an anti-menem coalition of the UCR and FREPASO (the Alianza) placed first in the 1999 presidential and congressional elections. The Alianza s victory ended a decade of PJ dominance and seemed to solidify FREPASO s status as Argentina s third major party. At the height of FREPASO s electoral success, several scholars assumed that FREPASO would become institutionalized (Abal-Medina 1998; Novaro and Palermo 1998) and contribute to the normalization of Argentina s historically unstable democracy (Levitsky 2000). 1 Abal Medina (1998a) elaborates this point. 288

297 Yet in 2001, FREPASO collapsed, never to be rebuilt. The immediate cause of the collapse was a congressional electoral crisis precipitated by economic calamity. The Alianza s first year in government coincided with one of the worst economic downturns in Argentine history. The downturn, combined with a 2000 Senate corruption scandal, seriously damaged the Alianza s public image and led FREPASO s leader, Carlos Chacho Álvarez, to resign the vice presidency in October of Months later, in the 2001 midterm elections, FREPASO suffered a major setback, losing sixty percent of its congressional seats. Shortly thereafter, the party dissolved. Figure 6.1: FREPASO in the Chamber of Deputies Paradoxically, however, the root cause of FREPASO s failure lay in the party s unencumbered access to media and resulting organizational weakness. After breaking from the PJ in 1990, FREPASO s founders contested six separate national elections in less than a decade. Organization-building, by consuming scarce time and resources and placing procedural and 289